Spatio-Temporal Boundaries of African Civilizations Reconsidered David Wilkinson University of California, Los Angeles Presented to the 22nd Annual Meeting of the International Society for the Comparative Study of Civilizations, Scranton, Pennsylvania, June 3-6, 1993 Abstract During a previous round of collective efforts by civilizationists to provide rosters of civilizations, a definition based on level (cities) plus autonomous politico-military interaction-networks passed two (Egyptian/Northeast African, later fused into Central, and West African) candidates from the continent of Africa. Further historical accounts and urban data collections justify adding two more: East African (Coastal/Swahili) and West Central African (Kongo/Tio). A third is probably verifiable: African Great Lakes. Several other candidates have incomplete cases; decisive evidence in their cases, as for African Great Lakes, must come from systematic archaeology. Spatio-Temporal Boundaries of African Civilizations Reconsidered How many civilizations have there been, where, when, by what criteria, on what evidence, with what origins and destinations? This question, which is essential and primary to the comparative study of civilizations, has been variously answered, and answers have evolved with theory, definitions and evidence. (See notably e.g. Wescott, 1968, and Melko and Scott, eds., 1987) While such states as the United States, Colombia and Ireland, and such areas as Central Asia, Siberia, Micronesia (Pohnpe) and Polynesia contain intriguing and controversial candidates and potential candidates, the largest number of such claimants are to be found in a single continent - Africa. Africa has also been the subject of much historical attention, controversy and revision in the last two decades. New sources of data worldwide have emerged, and old ones have been improved, in the last decade or so, with relevance to the questions of the rosters, duration and extent of civilizations. This paper accordingly re-addresses itself to the question: How many civilizations were there in Africa? When? Where? The definition here applied to new collections of evidence remains unchanged from previous rosters. Two criteria are basic. (1) A civilization is a society of a certain character, i.e. an urban society, a society with cities. Cities are operationally defined as population centers with not fewer than 104 (10,000) inhabitants. Smaller conurbations will be labeled "towns"; no upper bound is here defined. Cities are usually (but not necessarily or by definition) sexually integrated (monastic conurbations qualify, but are infrequent) and of fixed location (but Sarai and Kibuga qualify). Periodic or occasional gatherings (Woodstock, perhaps Metalanim) do not qualify. "Cities" and "civilization" are etymologically related. Webster's New International Dictionary (2nd edition) shows how both stem from the Latin civis (citizen), one by way of civitas (citizenship; city) and French cite (city), the other more directly. The Oxford English Dictionary (2nd edition) connects them thus: "civilization" = "3. civilized condition or state; a developed or advanced state of human society"; "civilized" = "1. Made civil, in a state of civilization"; "civil" = "I. 1. Of or belonging to citizens; consisting of citizens, or men dwelling together in a community...."; and "citizen" = "1. An inhabitant of a city or (often) of a town...." "Cities" and "civilization" are also customarily related in civilizational theory. Bagby (1963:162-163) defined civilization as "the kind of culture found in cities." Quigley (1961:31,76) made cities a preliminary and partial defining property of civilizations (which all his listed "civilizations" in fact possessed). Spengler (1926-1928:II, 87-110) made cities--which however he defined by "soul" and only incidentally by size--and urban succession an essential feature of civilization. Only Toynbee struggled to avoid the connection between city and civilization; he pressed the claims (1961:276-277) of a "Nomadic" civilization to be viewed as having existed, at least in an "arrested" (1961:547) or "satellite" (1961:553) status. But no such "Nomadic" civilization appears in Toynbee's final roster of civilizations, not even among the "satellites" (1961:560-561); and every society finally cited by Toynbee as a "Full-blown" and "Independent" civilization (1961:558-561) can be shown to have had cities (see e.g. maps to Wilkinson, 1992-1993). The number 10,000 as a threshold separating "town" from "city," while not fully binding, seems warranted by usage and theory. Chandler (1987) enters no cities with populations smaller than 10,000 in his lists. And Iberall and Cardon (1973:38-40) have found a discontinuity in the size distribution of societies described in Murdock's Ethnographic Atlas (1967) which suggests that between society-sizes of about 2000 and 7000 persons there is an instability boundary, cusp, or bifurcation locus that separates two different sorts of societies. One, with tens and hundreds of members, is characteristically an isolated monoculture with small- organization stability that expands demographically by fission and diffusion through space at basically constant density (Iberall and Wilkinson 1984a), culturally homogenous within and varying slightly across immediate boundaries. The other, with thousands and tens of thousands, is characteristically a condensation point in an above- normal density population field whose expansion has been globally constrained (Iberall and Wilkinson 1986) as perhaps also locally (Edwards 1987) with remixing of ethnicity (Iberall and Wilkinson 1984b), and by reason of the remixing and fusion of lineages is characteristically a polyculture (Iberall and Wilkinson 1993). Cities are then herein the first defining criterion of a civilization, or rather of "civilization." The second criterion, which allows us to distinguish a plurality of societies as "civilizations," is that of internal politico-military interaction, and external politico-military isolation. A civilization is that urban society which has its own distinct politico-military history, a system of states (one or many) with their own wars and diplomacy. This definition noticeably omits economic and cultural criteria. A society bound by a trade network is defined as an oikumene; it may or may not be coextensive with a civilization; today there is one global oikumene coextensive with one global civilization; in previous ages oikumenes have usually not been coextensive with civilizations (Wilkinson, 1992-1993). Civilizations are also distinct from cultures and culture- areas. Civilizations are characteristically polycultures. (Iberall and Wilkinson, 1993). This premise radically distinguishes the present argument from that of Spengler, for whom civilizations and cultures were spatially coextensive (but temporally distinct - a civilization was a dead or ossified culture, 1926-28:I,31). It less radically distinguishes it from that of Toynbee, for whom monoculture and polyculture were successive (and declining) phases of a civilization - the nearest concept to polyculture in Toynbee is some combination of "Schism in the body social, which, with "Schism in the soul," is an essential feature in "the process of the disintegration of civilizations." (1939:part V) It moderately distinguishes it from that of Quigley, for whom polyculture (stage of "mixture") was an inchoate period in the creation of a civilization which would be monocultural or at least have a dominant, integrated culture after its earliest stages, once the society was pervaded by a single "instrument of expansion" (roughly, socioeconomic form:69-80). It avoids Sorokin's critiques of Toynbee (and other civilizationists) for the procrustean imposition of cultural unity iupon what were historically "congeries" (Sorokin, 1950:213 et passim; see also Wilkinson, 1994). Rather than being integrated, functionally differentiated macrocultures, civilizations are social fields in which cultures appear, expand, move, change, fragment, syncretize, conflict, turn over, vanish. * * * What African societies then possessed cities over 10,000 population, and politico-militarily autonomous histories? Previous papers in this series (1987a, 1987b) found two African civilizations demonstrable--Egyptian (later, Central) and West African. Nine more--Ethiopian, Nubian, Sudanic, Zimbabwe I, Zimbabwe II, East African, Ugandan, Kongo, and Central African-- were considered to be possible but unproven. (1987b:36) At that time, a paper by Gordon Hewes (1987), likewise derived from a series of "Boundaries" discussions among civilizationists, defended a considerably larger roster of ten Subsaharan civilizations: 1. Nubian; 2. Saharan Berber; 3. Ethiopic; 4. Sudanic--i. Western (e.g. old Ghana and Mali, Songhai, Wolof, Mossi, Hausa, Fulani), ii. Central (e.g. Kanem-Bornu, Bagirmi, Darfur), iii. Guinea Coast (e.g. Asante, Dahomey, Yoruba, Benin, Warri); 5. Bantu--i. (Monomatapa-Changamire-Luba-Chokwe-Kongo-Kuba), ii. Lakes (Kitwara, Bunyoro, Buganda, Burundi), iii. Swahili; and 6. Madagascar. (1987:91-93) Recently the discussion of the roster of African civilizations has been reopened. A paper presented by Gloria Thomas-Emeagwali (1992) proposes or reasserts, in addition to Egyptian/Nubian, the following candidate Black African civilizations: Axumite and Monophysite/Ethiopian; Western and Central Sudanic (Ghana/Mali/ Songhai/Kanem-Borno); Central African/Kongo; perhaps a coastal Swahili/East African; Zambian/Zimbabwe/Monomatapan/Southern African; a Lacustrine East African; and a Niger/Benue (Oyo/Nupe/Kwararafa). Furthermore, in the interval since the Boundaries discussions (which, though published in 1987, took place around 1979) and rosters, several data sources have become available. 1. The UNESCO General History of Africa has published the volumes edited by Mokhtar (1981), Niane (1984), Elfasi and Hrbek (1988), and - as this paper was being written - Ogot (1992). 2. The Cambridge History of Africa has added the volumes edited by Clark (1982) and Oliver and Sanderson (1985). 3. The urban data collection of Chandler (1987) was published. 4. Less important except illustratively, but crucial in that respect, a variety of atlases of African history - McEvedy (1980), Kwamena-poh et.al. (1982), Ajayi and Crowder (1985) - have appeared. A reappraisal is timely. * * * The following 18 candidate civilizations will be evaluated, as will emergent candidates that are combinations of entities on the list. The order is partly chronological, partly regional. 1. Egyptian, later Central 2. Nubian 3. Ethiopian 4. Kaffan 5. Northwest African (Punic) 6. West African (Western Sudanic: Ghana/Mali/Songhay/Senegambian) 7. Chadland (Kanem/Bornu) 8. Hausaland 9. Lower Niger (Nupe/Yoruba/Edo) 10. Dahomeyan 11. Gold Coast 12. South Central African (Zimbabwe/Momomatapa) 13. East African (Coastal/Swahili) 14. West Central African (Kongo/Ndongo) 15. Central African (Shaba; Lunda/Luba) 16. African Great Lakes 17. Madagascan 18. South African In accordance with the city-dependent definition of "civilization," we must ask, when were there cities in these areas? A suitable approach to this question is to examine the most comprehensive and ambitious urban data collection extant, Tertius Chandler's lists from 4000 Years of Urban Growth (1987). These must be employed critically (see Wilkinson, 1992-1993); but they allow a substantial entree into the subject, and represent an independent data source, the best available, collected on consistent comparative principles. Let us examine Chandler's list of African cities (updated by personal communication, 1992). African Cities DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) 2250 BC Memphis 1987:460 Heliopolis Heracleopolis to c. 30,000 2000 BC Memphis 60 1987:460 Thebes Heliopolis to. c. 25,000 1800 BC Thebes 1987:460 Memphis Heliopolis Kerma (Nubia) Heracleopolis Asyut to c. 20,000 DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) 1600 BC Avaris 100 1987:460 Memphis Kerma Nekhen (=Hierakonpolis) to c. 24,000 1360 BC Thebes 80 1987:460 Memphis 32 Amarna 30 Heliopolis 30 to c. 24,000 Damietta 24-l7 4/25/92 letter Bubastis 17-15 Heracleopolis 15-10 Hierakonpolis 15-10 Tanis 15-10 Lisht 15-10 to 10,000 DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) 1200 BC Memphis 50 1987:460 Thebes 40-30 Heliopolis Tanis to c. 24,000 Bubastis 24-15 4/25/92 letter Sebennytus 24-15 Buto 15-12 Damietta 15-12 Hierakonpolis 15-12 Ashmunein c. 12 Syene c. 12 to c. 12,000 1000 BC Thebes 50 or more 1987:460 Memphis 50-25 Heliopolis to c. 25,000 800 BC Thebes 50 or more 1987:460 Memphis 50-36 Heliopolis to c. 25,000 DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) 650 BC Memphis 70-60 1987:460 Sais 48 Napata (Nubia) 45-40 to 30,000 Meroe (Nubia) 30-25 4/25/92 letter Heliopolis 30-25 Avaris 25-20 Buto 25-20 Athribis 25-20 to 20,000 430 BC Memphis 100 1987:461 Carthage (Tunisia) 50 Meroe 49-45 Cyrene (Cyrenaica) 35 to 30,000 Heliopolis 24-20 4/25/92 letter to 18,000 200 BC Alexandria 300 1987:462 Carthage 150 Memphis 70-60 Cirta (= Constantine, Algeria) 40-38 Meroe 38-35 to 30,000 Jol (Cherchel, Numidia) Oxyrhynchus 30-24 4/25/92 letter to 20,000 DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) AD 100 Alexandria 250 1987:463 Carthage 100 Memphis 50-45 Oxyrhynchus 34 to 30,000 Meroe 30 4/25/92 letter Cirta 30 Tingis (= Tangier) 28-27 Cyrene 25 Arsinoe (Crocodilopolis, Egypt 24 Volubilis (Morocco) 20 Rusicada (Algeria) 20 Hadrumetum (= Sousse, Tunisia) 20 to 20,000 AD 361 Alexandria 125 1987:464 Carthage 66 Leptis (Libya) 50 Axum (Ethiopia) 50-45 Hippo Regius (= Bone, Algeria) 40 to 40,000 Memphis 40-35 4/25/92 letter Oxyrhynchus 34 Qostol (S. Egypt) 34-30 Hadrumetum 30-25 Thysdrus (Tunisia) 30-25 Arsinoe 24 Tingis 24-21 Timgad (= Thamugadi, Algeria) 21 Volubilis 20 to 20,000 DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) AD 500 Carthage 100 1987:465 Alexandria 100 Coptos 50-40 to 40,000 Memphis 35-50 4/25-92 letter Leptis 35-30 Hadrumetum 30-25 Hippo 24-20 to 20,000 AD 622 Alexandria 94 1987:466 Carthage 50 Coptos 45-40 to 40,000 Memphis 35-50 4/25/92 letter Zimbabwe 30 DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) AD 800 Fostat 100 1987:55 Alexandria 95 Kairwan (Tunisia) 80 Gao 72 Qus (Egypt) 40 to 40,000 Ghana 30 4/25/92 letter to 30,000 Sijilmessa (Morocco) 24 1987:55 to 24,000 "cities" (no size estimate) 1987:55 Damietta Tunis Tinnis Tahert (Algeria) Meknes (Morocco) Dongola Axum Awdaghost (Ghana) Zimbabwe AD 900 Alexandria 175 1987:468 Fostat 150 Kairwan 100 Qus 40 to 40,000 Gao 40-36 4/25/92 letter Sijilmessa 35-32 Tahert 35-32 to 30,000 DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) AD 1000 Cairo 135 1987:55 Tinnis 83 Kairwan 80 Qus 45 to 40,000 Manan (= Matan, Bornu) 40-35 4/25/92 letter Mahdia (Zirids, Tunisia) 40-35 Meknes 40-35 Fez 30 Ghana 30 Soba (Aloa) 30 to 30,000 Dongola 25 1987:55 to 25,000 "cities" (no size estimate) 1987:55 Damietta Tlemcen Constantine Sijilmessa Sale (Morocco) Tademekka (Ghana) Zimbabwe DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) AD 1100 Marrakesh (Morocco) 150 1987:470 Cairo 150 Fez 125 Tinnis 110 Bougie (Algeria) 50 Qus 50 Meknes 40 to 40,000 Gao 35-32 4/25/92 letter Mahdia 35-32 Dongola 30 to 30,000 DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) AD 1200 Fez 200 1987:55 Cairo 200 Marrakesh 150 Damietta 100 Qus 50 Rabat (Morocco) 50 Alexandria 50 Bougie 50 Ceuta (Morocco) 40 to 40,000 Tlemcen 35-33 4/25/92 letter Meknes 33-32 Tunis 30 Dongola 30 to 30,000 Soba 25 1987:55 Gao 25+ Ghana 25 Kano 25 Njimiye 25-20 Walata 25-20 Zagha (Tekrur) 20 Kilwa 20 to 20,000 "cities" (no size estimate) 1987:55 Constantine Kairwan Mahdia Sijilmessa Lasta (Ethiopia) Bussa (Borgu) Ngala (So) Zimbabwe DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) AD 1300 Cairo 400 1987:474 Fez 150 Damietta 108 Marrakesh 75 Alexandria 65 Tunis 50 Tlemcen 50 Qus 50 Bougie 50 Meknes 40 Ceuta 40 Sale 40 Njimiye 40+ Mali 40 to 40,000 Asyut 30 1987:56 Dongola 30 Turunku 40-30 Kano 30 Gao 25 Zimbabwe 25 Walata 25-20 Soba 20 Kilwa 20 Jenne 20 to 20,000 "cities" (no size estimate) 1987:56 Mansura (Egypt) Mahdia Constantine Oran Tegulat (Ethiopia) Ife (Yorubas) Ngala Nufi (Nupe) Bussa DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) AD 1400 Cairo 360 1987:476 Fez 125 Damietta 90 Tlemcen 70 Tunis 50 Marrakesh 50 Mali 50 Oyo (Yorubas) 50 Qus 50 to 45,000 Alexandria 40 1987:56 Bougie 40 Gao 40 Meknes 35 Sale 35 Ambessi (=Mbanza Kongo) 35+ Zimbabwe 35 Nupe 50-30 Ceuta 30 Constantine 30 Axum 30 Kilwa 30 Kano 30 Dongola 25 Turunku 30-20 Durbi 30-20 Rao (Jolof, Senegal) 30-20 Timbuktu 20 Jenne 20 to 20,000 "cities" (no size estimate) 1987:56 Asyut Mansura Mahdia Oran Azammur Krenik (So) Ife Benin Ouagadougou (Mossi) DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) AD 1500 Cairo 400 1987:478 Fez 130 Tunis 65 Gao 60 Oyo 60 Kano 50 Marrakesh 50 to 50,000 Bougie 40 1987:57 Constantine 40 Tagust 40 Kazargamu (Bornu) 40 Mali 40 Sao Salvador (Congo) 40 Tlemcen 35 Axum 33 Turunku (Zegzeg) 40-30 Meknes 30 Oran 30 Kilwa 30 Gobir 28 Mahdia 30-25 Taza 25 Timbuktu 25 Azammur 24 Agades 25-20 Katsina 25-20 Rao 25-20 Ngala 25-20 Algiers 20 Tedsi 20 Dongola 20 Jenne 20 to 20,000 "cities" (no size estimate) Rosetta 1987:57 Damietta Asyut Chonga (Kaffa) Nupe Ife Benin Ougadougou Ijebu (Yorubas) Chitako (Monomotapa) AD 1600 Cairo 200 1987:57 Marrakesh 125 Fez 100 Algiers 75 Kazargamu 60 Zaria 60 Katsina 60 Oyo 60 Tunis 50 Benin 50 Kano 40 Surame (Kebbi) 40 Sennar (Nubia/Sudan) 40 Gobir 35 Gbara (Nupe) 33 Constantine 30 Tagust 30 Agades 30 Dongo (Angola) 30 Loanda (Angola) 30 Damietta 30-25 Tlemcen 25 Timbuktu 25 Masenya (Bagirmi) 20 "cities" (no size estimate) Bulaq (near Cairo) 1987:57 Meknes Bonga (Kaffa) Ife Kikiwhary (Assin, Gold Coast) Chitako Loango DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) AD 1700 Cairo 175 1987:58 Algiers 85 Fez 80 Gondar (Ethiopia) 72 Meknes 70 Tunis 69 Katsina 60 Sale-Rabat 50 Kazargamu 50 Zaria 50 Oyo 50 Agades 45 Kano 35 Jima (Nupe) 35 Constantine 30 Bonga 30 Sennar 30 Masenya 30 Allada 30 Naya (= Chibiri, Gobir) 30 Dongo 30 Loango 30 Marrakesh 25 Benin 25 "cities" (no size estimate) Damietta 1987:58 Bulaq Emfras (Ethiopia) Kebbi Zamfara Puje (Kwararafa/Jukun) Ife Mengo Lunda's capital DATES and THRESHOLDS SIZE SOURCE in CHANDLER (Thousands) AD 1800 Cairo 186 1987:59 Tunis 90 Oyo 80 Algiers 73 Katsina 70 Kazargamu 50 Alkalawa (Gobir) 50 Meknes 45 Rabat-Sale 43 Fez 40 Zaria 40 Kumasi (Ashanti) 40 Ilorin 40-30 Gbogun (Yorubas) 40-30 Ife 40-30 Ogbomosho (Yorubas) 40-30 Marrakesh 30 Segu (Bambara) 30 Bonga 30 Masenya 30 Kebbi 30 Kano 30 Kiama 30 Asyut 25 Constantine 25 Mouzangaye (Sakalavas, Madagascar) 25 Yendi (Dagomba) 30-24 Bulaq 24 Abomey (Dahomey) 24 Gbara 24 Damietta 23 Tripoli 20 Kairwan 20 to 24,000 "cities" (no size estimate) Ankober (Ethiopia) 1987:59 Kiawa (Zamfara) Sangha Mengo Locating Chandler's African cities; some revisions to the list. Figure 1 shows the locations of these cities, and of the 18 candidate civilizations. It embodies some modifications, both [FIGURE 1 GOES ABOUT HERE] as to city names and as to locations, by comparison to Chandler's maps (1987:60-68) and lists, as follows: Alkalawa. Location per Hogben and Kirk-Greene, 1966:117. Awdaghost. Location per Rand McNally, 1992:64. Chitako. Location per Ransford, 1968:28. Chonga. Kaffa's capital c. AD 1500 is so labeled in Gruehl, 1932:177; it is called "Schonga," which should be transliterated "Shonga," in Gruehl, 1938: map. But this point is moot, inasmuch as Bieber, 1923:517-522, provides a more detailed dynastic history of Kaffa than Gruehl, 1932:176-178, and Bieber places the court of Kaffa at Shonga only c.1425-1460, at Addio c. 1460-1495, and at "Schadda," i.e. "Shadda," c. 1495-1530. Hence in the 1500 list one should substitute Shadda, q.v. Durbi. Location per Smith, 1972:192 and Palmer, 1936:129. Emfras. Location per Bruce, 1790: map after errata. Gbara. Location per Nadel, 1942: map after 420. Gbogun. Location a compromise between Mabogunje, 1968:77 ("Gbongon") and Law, 1977:279. Ghana. Location used is Kumbi Saleh, per Rand McNally, 1992:64, vs. Awdaghost or Walata (cf. Chandler, 1987:55, 294). Gobir. As "Bir Alele," Palmer 1936:84. As "Birnin Lalle" in Gulbin Tarka c. 1450, Laya, 1992:461-462. Location relative to Naya/Tibiri per Last, 1985:215. Location of the Tarka valley per The World Atlas (1967:168). Jima. Location per Nadel, 1942: map after 420. Kiawa. Location per Hogben and Kirk-Greene, 1966:117. Kikiwhary. Location per Dupuis, 1824:35-37 and map after Part I:264. The site of a "very extensive city, which was destroyed some ages back by an irruption of the Dagomba." Chandler dates its city status AD 1600 (1987:57). The history of the Dagomba (Levtzion, 1975:187-189; Izard and Ki-Zerbo, 1992:339-340) does not place them this far south (Kikiwhary was in Assin). Perhaps Dupuis' informant meant rather "the Denkyira," who did conquer Assin in the late 17th century (Boahen, 1992:415) and waged a punitive war there in 1697 (Rodney, 1975:300). Krenik. (Debabe Ngaya) Location per Barth, 1965, vol. II: map after p. 425. Lunda's capital. A city in 1700, per Chandler, 1987:58. In Gray, ed., 1975:326,330, it is named "Musumba," but Nziem (1992:602) notes that "musumba" means "capital," and Birmingham (1975:271) notes that there were "several musumbas" in the nineteenth century. This paper uses the name employed by Vansina (in Ogot, ed., 1992:604), "Rund," q.v. Majunga. Replaces Mouzangaye, q.v. Location per The World Atlas (1967:174). Now often Mahajanga. Mali. Location (as "Niani") per Niane, 1984:128, 136; and The World Atlas (1967:168). Manan. See Njimiye. Mengo. See Chandler, 1987:68; as "Mengo area," 1987:300. Mengo hill was the ultimate vicinity of the mobile kibuga, or capital of Buganda: Gutkind, 1963:9-10. But the name of the capital is given by Alpers (1974:244-245) as "Rubaga," q.v., which accordingly replaces Mengo on the accompanying maps. Mouzangaye. Sakalava capital, Chandler 1987:301; cf. 1987:68. So identified by Malte-Brun (1827:99). Vivien (1879-95: "Madsanga") equates Mazangaye, Madsanga, Madjanga, Mojanga and Majunga. Most current sources (e.g. Ajayi and Crowder, 1985:51) use Majunga, q.v., and it is accordingly substituted. Naya. Also Chibiri, Tsibiri, or (on most current maps) Tibiri. Identification and location per Chandler, 1987:296; Adeleye, 1972:513; Last, 1985:215; and Times Atlas of the World, 1990: Pl. 90. Ngala. Location per Mabogunje, 1968:70. Njimiye. Either Manan has been renamed Njimiye over time, and the original Njimiye site is somewhere east of Lake Chad, or it has not, and Manan is lost somewhere to the northwest of present (and past) Njimiye (Nguigmi). For the first theory, see e.g. Smith, 1972:159, 169-170; maps consistent with it in Ajayi and Crowder, 1985:24, and Lange, 1984:257. The maps herewith included embody the second theory, and are consistent with Chandler (1987:61-63) and Freeman-Grenville (1991:45). Nufi. Chandler (1987:63-65) gives this a location 1300-1500 in the vicinity of later Gbara and Jima (q.v.). His 1400 and 1500 lists however call it Nupe (1987:56-57), and they are identified as the same 1987:296. I agree with the identification, but prefer another location: see Nupe. Nupe. Location for 1400 in the vicinity of Nku and Nupeko on the Kaduna S of Gbara, those sites per Nadel, 1942:map after 420. Nadel places the story of Nupe's legendary founder Tsoede about then despite later versions (1942:75). Nadel accordingly much doubts (1942:404) that Ibn Batuta's reference (AD 1353) to Youfi/Noufi applies to Nupe (even though that was also called Nouffie); but there seems no better alternative. Oyo. Old Oyo; location per Mabogunje, 1968:77. Puje. Location per Hogben and Kirk-Greene, 1966:196, 117. Rao. Chandler cites this as perhaps "Thieung" and probably the Jolof capital 1212-c. 1566 (1987:288), and shows it (1987:64- 65) in 1400 and 1500 as just north of the mouth of the Senegal river. Davies (1967:272) speculates that the mound burials here reflect the Jolof capital. But the Atlas National du Senegal sees "Walo" as only its first brief site, and locates the real Jolof capitals inland at "Tyeng" and Walkokh near present Linguere. Walkokh (q.v.) is here used. Rubaga. Replaces Mengo, q.v. Location per Kinambo, 1989:253. Rund. Lunda's capital, q.v. Location per Vansina in Ogot, ed., 1992:604, and Times Atlas of the World, 1990: Pl. 91. Salgha. Vs. Sangha, q.v. In "Entaa" (Dupuis, 1824:I,170); "the chief city of Ghunja" (1824:II,cxxv) which includes the Ashantee and Dahomey empires (1824:I,lxxxiv); "the metropolis of Ghunja" (1824:I,xxxvi). Reported to be "of twice the size of Coomassy," about 400,000 (1824:II,xl), which Chandler doubts (1987:297), I think correctly as Dupuis makes no eyewitness report on the demographics of either city; said to be tributary to "Ashantee" but also on the Dahoman frontier (1824:II,xxxix). Location of Salgha from map after 1824:I,264 shows it to be same as modern Salaga. Bonnat (Johnson, 1965:SAL/1/2) estimated the former probable population of "Salaga" at 40,000, its current (i.e. in 1876) population at 12 to 15,000. V.S. Gouldsbury (Johnson, 1965:SAL/11/1) in the same year estimated the current population at 8,000, with a prior floating population of 10,000. Dupuis, or an informant, seems to have added a zero to his estimate. Sangha. In Chandler, 1987:59, 68, 297; but this must be Salgha, q.v. Shadda. Replaces Chonga, q.v. Location per O. Bieber, 1948:111 (as "Schadda"). Surame. Location per Mabogunje, 1968:51. Tagust. Location (as Tagavost) per Leo Africanus, 1956: map after p. 70. Tedsi. Location per Leo Africanus, 1956: map after p. 70. Tegulat. Location per Vivien (1879-95: "Tegulet"). Tinnis. Location per Oliver, ed., 1977:20, in contrast to Fage, ed., 1978:498, which would place it at Tanis. Turunku. Location per Hogben and Kirk-Greene, 1966:216, 117. Walata. Location per Rand McNally, 1992:64. Walkokh. Substituted for Rao (q.v.) as Jolof capital. Location per Atlas National du Senegal (1977:52-53). Zamfara. Location per Mabogunje, 1968:51. * * * The next step is to classify Chandler's African cities, as revised, by candidate civilization (omitting Egypt, which is a given), and examine the main comprehensive sources (The Cambridge History of Africa and the General History of Africa) for supplementary urban data and for politico-military interaction data. As will be seen, all the candidate civilizations listed (with the exception of South African, included for reasons to be discussed), and no others, had in their areas cities on Chandler's lists at some time between 2250 BC and AD 1800. Some had them often (e.g. Nubia), some briefly (e.g. Madagascar). 1. EGYPT For the purposes of this study, an Egyptian (or Northeast African) civilization is taken as given. Only its spatial boundaries will be examined, and the examination will concern, and be carried out in the sections devoted to, the other entities which are candidates for civilizational status to the extent that they were not semiperipheries of Egypt (and, later, of the Central civilization which grew out of the fusion of Egyptian with Mesopotamian civilization c. 1500 BC--Wilkinson, 1987b). Egypt had cities on all of Chandler's list from first to last, usually possessing the largest city in Africa. As to Egypt's startup date as a civilization, the epoch of the Gerzean city-states like Abydos, wicked/defeated god Seth's Naqada, and glorious/victorious god Horus's Hierakonpolis struggling for divine glory and the eastern-desert gold trade in some inestimable balance, mid- to late-4th millennium BC, still seems most plausible. (Trigger, 1982:517, 512-513, 526, 483) 2. NUBIA Nubian cities. The following cities of Nubia are given size range estimates on Chandler's lists for the years cited (historians' comments are included where and as relevant): Size Lower bound of list 2250 BC: none c. 30,000 2000 BC: none c. 25,000 1800 BC: Kerma none c. 20,000 Kerma "seems to have been the most important urban centre in the Kingdom of Kush whose name appears in Pharaonic texts from - 2000." (Adam, 1981:239) 1600 BC: Kerma none c. 24,000 1360 BC: none c. 10,000 1200 BC: none c. 12,000 1000 BC: none c. 25,000 800 BC: none c. 25,000 650 BC: Napata 45-40,000 Meroe 30-25,000 20,000 430 BC: Meroe 49-45,000 18,000 200 BC: Meroe 38-35,000 20,000 AD 100: Meroe 30,000 20,000 AD 361: none 20,000 AD 500: none 20,000 AD 622: none 24,000 AD 800: Dongola none none AD 900: none 30,000 AD 1000: Soba 30,000 Dongola 25,000 none AD 1100: Dongola 30,000 30,000 AD 1200: Dongola 30,000 Soba 25,000 none AD 1300: Dongola 30,000 Soba 20,000 none AD 1400: Dongola 25,000 none AD 1500: Dongola 20,000 none Shinnie dates the destruction of Old Dongola arouns 1365 (1978b:584-588). Kropacek believes Soba was destroyed by Abdallabi Arabs (1984:406-407), whose conquest of Aloa is placed by Hasan and Ogot in the late 15th century (1992:172-173). Kropacek finds it probable that "a considerable part of the earlier sedentary population turned nomad or semi-nomad in the dark period, as the margin of cultivation shrank." (1984:415) Perhaps the list should show Soba rather than Dongola AD 1400, and no cities AD 1500. AD 1600 Sennar 40,000 none Holt characterizes Sennar (of the Funj) and Qarri and Halfaya (of the Abdallab) as "permanent towns" or "small urban centers" which were the residences of the ruling families. (Holt, 1975:42) The Funj and Abdallab histories, as Holt narrates them, indeed more pastoral than urban. AD 1700 Sennar 30,000 none AD 1800 none none Sennar decayed as a trade center due to local violent rivalry with Darfur and the Shilluk kingdom, and was displaced between 1700 and 1775 by Shendi to its north. (Hasan and Ogot, 1992:183) The "town" of Shendi may have substituted for Sennar, though its caravans in 1814 were rather infrequent. (Holt, 1975:49-50) The town of Arbaji was sacked in 1783-84, as was Sennar itself in 1788- 89. (Holt, 1975:41-48) Mohammed Ali's troops in 1821 "found the fabled capital of the Funj to be little more than a heap of ruins." (Holt, 1975:48) By the time of the Turkish conquest Sennar may have sunk to 13-14,000 people. (Crawford, 1951:277.) There were then one or two cities at least in Nubia over most of the long period under study. Given the varying lower bounds, there may have been some cities of 10,000 or more throughout the period. Nubia and Egypt, late 4th to late 3rd millennium BC. An Egyptian military expedition reached the Second Cataract c. 3200 BC. (Adam, 1981:234) "An ebony tablet from the time of Hor-aha, the first king of the Egyptian first dynasty, seems to celebrate a victory over Nubia." (Sherif, 1981: 248) Zayed (1981:141) notes that "Already in the first dynasty forts protected the south of Egypt against its southern neighbors" and that Nubians served as archers in the Egyptian army. Hostilities between Egypt and Nubia are recorded in the second dynasty; Nubia was subjugated in the fourth dynasty, colonized in the fifth, under a hegemonic forced peace in the sixth. (Sherif, 1981:248-251) Nubia was occupied by a new cattle-herding people ("C-group") during the First Intermediate Period, 2240-2150 BC, of Egyptian imperial collapse. (Sherif, 1981:251) Nubia and Egypt, late 3rd to mid-2nd millennium BC. The Middle Kingdom Pharaohs re-invaded Nubia in the eleventh dynasty and occupied it in the twelfth (1991-1786 BC), building fortresses. (Sherif, 1981:255-258) The twelfth dynasty Pharaohs engaged in military expeditions against the south. (Adam, 1981:239) The southern border fortresses continued to be occupied in most or all of the Second Intermediate Period. (Adam, 1981:240) Kush had political-military relations with Thebes and the Hyksos 1650 BC to 1580 BC. (Adam, 1981:241) Nubia and New Kingdom Egypt, late 2nd millennium BC. The New Kingdom Pharaohs campaigned repeatedly against Nubia, both conquering and depopulating it. (Adam, 1981:241) Tuthmose I (1530-1520 BC) conquered Nubia and ended the kingdom of Kush. (Sherif, 1981:261, 265) During the eighteenth dynasty Nubia occasionally revolted but usually paid tribute to Egypt. (Sherif, 1981:266-267) In the nineteenth dynasty Egypt exploited, built in, and acculturated Nubia. (Sherif, 1981:266-273) Nubia took part in the internal Egyptian struggles of the twentieth dynasty, and supplied a Pharaoh in 1085 BC. (Sherif, 1981:273-274) Nubia and Egypt, 11th to 1st centuries BC. There followed three centuries of chaos, the "Third Intermediate Period," 1085-715 BC. During this time a Kushite kingdom slowly expanded northwards. Under Shabako (715 BC) Kush supplied the hegemonic dynasty of Egypt, and led Egypt to repeated bouts of war with Assyria 701-663 BC, which ended only with the Assyrian conquest of Egypt. (O'Connor, 1982:879-892) In the "Late Period," the area between the second and fourth cataracts was under Kushite control from Napata, while that between the first and second cataracts was "a virtually uninhabited and contested area between the two powers" of Kush and Egypt. (O'Connor, 1982:901) An Egyptian raid into Kush in 591 BC, under Psammetichus II, captured Napata and motivated the Kushites to transfer their capital south to Meroe. (Leclant, 1981:285) The Persian king Cambyses sent spies to Meroe and then attempted an invasion in 525 BC, but was defeated by the desert. (Shinnie, 1978a:223) The Kushite state remained "approximately equal to Egypt in military and political strength throughout the Late and Ptolemaic periods." Relations varied from maintenance of the buffer zone with occasional open conflicts to the accord of Ptolemy IV that was followed by resettlement of the zone. (O'Connor, 1982:915-916) Nubia and Roman/Byzantine Egypt. In the Roman period, Meroe sacked Aswan. Petronius, prefect of Egypt, replied by capturing Napata in 23 BC. Queen Candace of Kush negotiated a peace, demilitarization and frontier demarcation treaty with Augustus in 21 or 20 BC. (Leclant, 1981:290. Shinnie, 1978a:246-247) Thereafter Kush had a period of some prosperity, although Nero's ambassadors c. AD 60 pronounced it too poor to be worth conquering. (Leclant, 1981:292) In AD 297, Diocletian withdrew the Roman frontier to the First Cataract, and invited "Nobatae" to settle Lower Nubia as subsidized allies against the (also subsidized!) enemy Blemmyes. (Law, 1978a:208-209) Somewhere before AD 330 the Meroitic kingdom collapsed and vanished. (Leclant, 1981:291-293; Shinnie, 1978a:260) Nobatae and Blemmyes remained in conflict in Nubia into the sixth century AD. It is not clear whether Nubia retained a city; if so it would perhaps be Faras. (Shinnie, 1978a:269-271) In the mid-sixth century, rival Melkite and Monophysite missionaries from Byzantium brought Christianity to Nubia, now organized as three states, Nobatia (capital Faras), Makuria (capital probably Old Dongola), and Aloa (capital Soba). The rulers of the three Nubian states seem to have been rivals, and the religious connections they established (Egyptian vs. Byzantine Greek) reflected political divisions within the Byzantine Empire, to which Egypt then belonged. (Shinnie, 1978b:559-564. Michalowski, 1981:331) Christian Nubia and Muslim Egypt, 7th-15th centuries AD. The Muslim Arabs conquered Egypt in AD 641. They then "signed with Nubia a treaty called a baqt" invoving a supply of slaves from Nubia for food, clothing and wine from Egypt. "During the seven centuries of Christian Nubia's independence, both sides regarded the treaty as valid in principle, but more than one armed clash occurred." (Michalowski, 1981:333) Shinnie provides a text (1978b:566-567) for the treaty, and cites "the resultant peaceful conditions which lasted, with a few war-like intervals, for about five hundred years." (1978b:567) Crown Prince Georgios of Nubia (Nobatia plus Makuria, capital Old Dongola, after c. AD 700) negotiated a revised treaty with the Caliph al-Mutasim whom he visited in AD 835. (Jakobielski, 1988:209) Nubia occupied much of Upper Egypt AD 962. (Jakobielski, 1988:211) Various diplomatic missions were exchanged between Nubia and Egypt in the 11th century. (Jakobielski, 1988:215) When in AD 1170 the Ayyubids replaced the Fatimids in Egypt, relations with Nubia became less friendly. (Shinnie, 1978b:583) Makuria may have allied with the last Fatimids; it fought and was defeated by the Ayyubids in 1172. The Ayyubids also drove Arab rebels into Makuria. (Kropacek, 1984:398-402) With the rise to power in Egypt of the Mamluks (1250-1517), Egypt began to fight with and intervene in Nubia from AD 1272. The Mamluk Baybars fought and defeated Dawud of Makuria in 1272-1276, and installed a Christian vassal on the throne. Disorder followed, to which the Mamluks replied by looting, slaving punitive expeditions which reduced Nubia's viability. There were periods of Nubian independence, rebellion, intervention, renewed dependence 1290-1324. (Kropacek, 1984:403-404. Shinnie, 1978b:584-588) Nubia was also undermined by nomadic Arab Beduin, whom the Mamluks drove into Nubia in 1302, 1351, 1353, 1378, 1395. (Kropacek, 1984:402) Dynastic troubles, calls for Mamluk assistance, an Arab destruction of Dongola, and the vanishing of Makuria after 1397 followed hard on these pressures. (Kropacek, 1984:403-404) Old Dongola was destroyed around 1365. (Shinnie, 1978b:584-588) Soba in Aloa was probably conquered and destroyed by Abdallabi Arabs before the Funj cattle-herder sultanate of Sennar was established. (Kropacek, 1984:406-407) The conquest of Aloa would date to the late fifteenth century. (Hasan and Ogot, 1992:172-173) Funj Nubia, Egypt and Ethiopia, 16th-19th centuries AD. Funj cattle nomads defeated and subjugated the Abdallabi in 1504, and this suzerainty was maintained, despite rebellions in the late 16th and mid-17th centuries, until the Turco-Egyptian conquest of 1820. (Hasan and Ogot, 1992:172-173) The Ottomans conquered Egypt in 1517. There were frontier clahes with the Funj in Nubia. Returning from an attack on Ethiopia and Portugal (i.e. the Portuguese in the Indian Ocean), the Ottomans conquered northern Nubia. They made futile attempts to expand in Nubia in 1554 and 1577; a third ateempt 1n 1622 had some success. (Hasan and Ogot, 1992:174-175) The Ottomans established a Red Sea base at Sawakin (Suakin), against the Portuguese, which traded with the Funj and may have fought them (or neighbors) in 1571. (Hasan and Ogot, 1992:175) Nubia was connected to Abyssinia by subordination (c. 1607), by fear of war (e.g. c. 1706, c. 1772) and by war (1618-1619, 1743- 1744), and by trade. (Crawford, 1951:180, 227-236, 251, 182-187, 239-242) The Funj extended their suzerainty westward into Kordofan from about 1554. (Hasan and Ogot, 1992:177) Further west of Nubia and east of Wadai, in Darfur, still-obscure kingdoms had formed: Daju in the 13th and 14th centuries; Tunjur c. 1400-1600. The Fur kingdom, less obscure, lasted c. 1640-1874. Darfur's strongest military and political connections seem to have been Chadian rather than Nubian, with spheres of influence, alliance, competition, subordination being developed in the 17th century. (Hasan and Ogot, 1992:187-189) Darfur moved eastward in the 18th century, conquering Kordofan from a Funj vassal late in the century and keeping it till the Turco-Egyptian conquest. (Hasan and Ogot, 1992:187-189) Late in the eighteenth century Funj collapsed into warring feudalities, weak, autonomous, local entities. (Crawford, 1951:260. Hasan and Ogot, 1992:185) In 1821 the Turks invaded and conquered Funj. (Crawford, 1951:262-275; Holt, 1975:48). Analysis. Except in the tables for 650 BC, 430 BC and 200 BC, the largest city size till AD 1600 is 30,000. At least one Egyptian city was always larger than the largest Nubian city, often several times as large. Was some constraining factor or process operating? S. Adam (1981:233) points to "the dual nature--military and economic--of the north-south contacts along the Nile valley." There may have been cities in Nubia throughout its history; or there may have been lapses during periods of disorder in Egypt, and after the fall of Aloa. Nubia, urbanized or nomadized, fighting or peaceful, weak or strong, dominant or dominated, assimilated or alienated, was continuously politico-militarily involved with Egypt, both during the Northeast African/Egyptian civilization and during the epoch of Central civilization since 1500 BC. When urbanized, Nubia seems to have been part of the Egyptian/Central semiperiphery except during the twenty-fifth dynasty, 7th century BC, when it was a core state and fought the dominant power of its time, Assyria, on a relatively equal footing. If urbanism ever lapsed, and before it rose and re-arose, Nubia was part of the Egyptian, then of the Central, periphery. Whether or not cities persisted, it seems clear that at no time do the records show an isolated, historically autonomous Nubian civilization. 3. ETHIOPIA Ethiopian cities. None appears through Chandler's AD 100 list. Thereafter, through 1800 (with some interspersed commentary): Size Lower limit AD 361 Axum 50-45,000 20,000 Were there cities earlier? Shinnie speaks of "sites" with stone building by the 5th century BC (1978a:262-264), de Contenson of "an embryonic form of urban life." (1981:354) The port of Adulis may have existed in the 3rd century BC, at the time of Ptolemy III (246-221 BC) as a trade emporium and a "town." (Shinnie 1978a:259- 265) De Contenson dates "the founding of the city of Axum" to the 2nd century BC (1981:341). Shinnie sees Axum as a "town" beginning to develop in the first century AD. (1978a:262-264) The Periplus of the Erythraean Sea (about AD 100) refers to Axum as a city, Koloe as a town, Adulis as a large village. Anfray also labels Axum a "city" and nine other sites "towns" or "market-towns." (Anfray, 1981:363-367) Shinnie sees Axum as "a town of some size" 3rd century AD, a "considerable power" mid-4th century AD, a royal capital with coinage, temples, palaces and stelae. (1978a:259-265) Chandler's dating is relatively conservative. AD 500 none 20,000 AD 622 none 24,000 AD 800 Axum none none The seventh and eighth centuries witnessed fighting with Arab Muslims, the destruction of Adulis, pressure from nomadic peoples, and a darkness. (Mekouria, 1988:560-565) AD 900 none 30,000 The center of the Ethiopian state had by the mid-9th century probably shifted south from Axum to "Kubar." (Tamrat, 1977:101- 102) AD 1000 none none Abyssinia suffered political decline in the late 10th century. (Tamrat, 1977:98-114) There was a sacking and a political and dynastic collapse in the second half of the tenth century. (Mekouria, 1988:565-568) AD 1100 none 30,000 AD 1200 Lasta none none Abyssinia began to expand again in the late 11th and 12th centiries. (Tamrat, 1977:98-114) A new capital was established at Adefa, in Lasta, about the middle of the 12th century, by the Zagwe dynasty, which continued to expand. (Tamrat, 1977:114) AD 1300 Tegulat none none AD 1400 Axum 30,000 none Zera-Yakob (1434-1468) settled the royal court at a new capital, Debre-Bihan in Shoa. (Tamrat, 1984:454) But cf. the next note. AD 1500 Axum 33,000 none AD 1600 none none The Abyssinian court was kept as an army in being, the vassal territories garrisoned--"this was the reason for the continuous mobility of the court and the absence of any large urban areas" during the 14th-early 16th centuries. (Tamrat, 1984:436-437) The mobile royal camp, however, "could easily approximate to the size of an average town." (Tamrat, 1984:437) AD 1700 Gondar 72,000 Emfras none none The imperial court remained a mobile tent-city until settling at Gondar in 1636. Muslim Harar was "the only genuinely ancient city on Ethiopian soil." (Haberland, 1992:705-712) AD 1800 Ankober none none Abyssinia disintegrated in the eighteenth century. (Haberland, 1992:732-733) There is usually an Ethiopian city, occasionally several, on the lists for the search period AD 361 - AD 1800. The high lower bounds do not exclude the possibility of small cities. But there are a number of political disintegrations and dark ages, which might be genuinely cityless, but invite further investigation. Ethiopia, 5th-1st centuries BC. Was Ethiopia part of the Central politico-military interaction network? 5th-4th century BC remains in Yeha, Melazo and Addi Galamo suggest both religious and political ties to Sabaea. (De Contenson, 1981:354) In the 3rd- 1st centuries BC it is "quite possible that the decline of Meroe, on the one hand, and the waning power of the South Arabian kingdoms, on the other hand, allowed the Ethiopians to control all trade in gold, incense, ivory, and products imported from the Indian Ocean." (De Contenson, 1981:359) Ethiopia, 1st-4th centuries AD. Were there Central politico- military links? Kobishanov affirms that Meroe and Axum were neighbors and allies, while Axum fought wars and fluctuatingly possessed or claimed hegemony in South Arabia. (Kobishanov, 1981:381) "From the close of the second century up to the beginning of the fourth, Aksum took part in the military and diplomatic struggle waged between the states of Southern Arabia.... In the fourth century they conquered the Meroe kingdom...." (Kobishanov, 1981:384) More cautiously: by AD 350 Axum was "a considerable power" and may have ended Meroe. (Shinnie, 1978a:259- 265) There was an Ethiopian religious connection with Egypt and an expedition into Nubia in the mid-4th century AD. (Shinnie, 1978a:260, 264-265) There may have been an Arabian conquest of Axum in the 3rd century AD. (Shinnie, 1978a:262-264) Ethiopia, 5th-6th centuries AD. Axum's policy and empire were of interest to the Byzantines. (Kobishanov, 1981:392) Byzantium in AD 524 had a formal treaty with Axum to fight Persia in Yemen. (Michalowski, 1981:329) In the sixth century AD (c. 525) Axum, perhaps with military aid from Byzantium, intervened in the Jewish- Christian-pagan struggles in South Arabia, leaving a garrison there. (Mekouria, 1981:412-415) "By the sixth century an Axum king was already appointing the South Arabian kings...." (Kobishanov, 1981:386) Ethiopia, 7th-15th centuries AD. There may have been warfare, and there were attempts to secure concord, between Nubia and Ethiopia c. AD 686-689. (Shinnie, 1978b:572) There was fighting with Arab Muslims in the seventh and eighth centuries. (Mekouria, 1988:560-565) Abyssinia expanded in the late 8th and early 9th centuries. (Tamrat, 1977:98-114) Arabs took control of the Dahlak islands, gradually establishing Islamic sultanates to the south of Ethiopia--Damut, Shoa, Ifat, Adal, Harar--which had political- military relations with Ethiopia from the 7th through the 16th centuries. (Cerulli, 1988) Egyptian Fatimids put pro-Muslim pressure on Ethiopia through the Coptic church in e.g. the 11th century. (Tamrat, 1977:105) In the late 13th century there was a system of states in Ethiopia: Christian Abyssinia, a Jewish "Falasha" state, Gojjam, Damot, and seven or so Muslim states--e.g. Shoa, Ifat, Dawaro, Adal, the Dahlak islands. (Tamrat, 1984:423- 425, 428) Damot was a dominant power in the 13th century. (Tamrat, 1984:431) Dahlak balanced between Egypt, Yemen and Abyssinia. (Tamrat, 1984:428) Abyssinia had political relations-- tribute, threat, armed conflict, annexation--with the Muslim states nearby, notably in the 14th-16th centuries. (Tamrat, 1977:140-177) The Solomonid dynast Amde-Siyon (1314-1344) conquered Hadya, Damot, Gojjam, Falasha, and made tributary the Muslim states Ifat, Dawaro, Sharkha and Bali. (Tamrat, 1984:434-435) Zera-Yakob, in the early 15th century, fought off attacks by Adal and rebellion by Hadya. His weaker and more fractious successors began to decline in face of Adal. (Tamrat, 1984:454) The fourteenth and early fifteenth- century Solomonids, unlike the more submissive late thirteenth- century and late fifteenth-century dynasts, were prepared to fight Egypt and to ally with Christian Europe in support of goals ranging from stopping the persecution of Egypt's Copts to ending Muslim supremacy in the Middle East. (Tamrat, 1984:450-453) Ethiopia, 16-17th centuries AD. In the first half of the sixteenth century the tide turned against Abyssinia. Adal, under the leadership of Ahmad Gran, and with recruits from the Islamic world, defeated and half-subjugated the Abyssinian empire. (Haberland, 1992:714-715) In 1541, Portugal, struggling to combat Turkish expansion and having sent embassies to procure an Ethiopian alliance already in 1487 and 1520-1526, provided assistance to Emperor Galawdewos (Claudius) in his successful restoration of the empire and destruction of the Muslim ststes. (Haberland, 1992:714-715) In 1541-1543 Portuguese and Turkish troops fought each other as part of a Christian-Muslim war in Ethiopia. (Tamrat, 1977:181-182) Oromo ("Galla") migrants now pressed the empire from the south, while the Turks pushed in from the Red Sea area of Eritrea. The Christians fought Ethiopian Jews. The empire revassalized the southwestern states of Inariya, Bosa and Kaffa (Gomar). Iberian missionaries implanted Catholicism, which was destroyed. Such were the events of the late sixteenth and early seventeenth centuries. (Haberland, 1992:714-732) A Funj king of Sennar, Nubia, was an Abyssinian vassal c. 1607; Abyssinia fought Nubia without much effect 1618-1619. (Crawford, 1951:180-187) In the second half of the seventeenth century Abyssinia reached an accommodation with the Oromo and the Turks, and had for the moment no enemies. (Haberland, 1992:732-733) Ethiopia, 18th-19th centuries. A French embassy tried to stir up a war between Funj and Abyssinia in 1706. (Crawford, 1951:227- 236) There was a second Funj-Abyssinian war 1743-1744, in which the Funj were victorious. (Crawford, 1951:239-242; Hasan and Ogot, 1992:178) Relations with Nubia were strained at the time of Bruce's journey in 1772 (Crawford, 1951:251); but Funj soon disintegrated. (Crawford, 1951:260) Abyssinia did so as well. (Haberland, 1992:732-733) Analysis. The contained size (except Axum AD 361) and number of cities are noteworthy. They are consistent with the general argument that to the extent that there were Ethiopian cities at any given era, Ethiopia was politico-militarily connected to Central civilization--via Yemen, via the Red Sea to Egypt, and/or via Nubia. There are no current grounds to posit a distinct Ethiopian civilization. Archaeological work would be required to reverse this judgment; the 5th and 4th centuries BC seem proper target times. 4. KAFFA Kaffan cities. Kaffa, source of coffee and of "coffee," a rainy forested highland area drained by the Omo river into Lake Rudolf/Turkana to its south, and currently part of Ethiopia, is some distance southwest of the main Ethiopian city-forming area. It was also a noted source of slaves, musk, gold, ivory, spices and honey. (Abir, 1968:55, 85-87; Zewde, 1991:16) It has no cities on Chandler's lists through AD 1400. Then: Size Lower limit AD 1500 Shadda (vs. Shonga) none none AD 1600 Bonga none none AD 1700 Bonga 30,000 none AD 1800 none none Other Kaffan 15th century capitals (Shonga, Addio) have already been mentioned; there was a 16th-century capital at Borreto about 1530-1565 (F. Bieber, 1923:518); Andaracha was the late 19th century capital (Zewde, 1991:66). Bonga may not have declined c. 1800: Bieber (1923:527) sees it at its height under Emperor Hotti Gaocho c. 1798-1821. Kaffan history. Kaffa was a Cushite state with a divine kingship and some Coptic Christian overlay, the best known of about ten such. Bieber dates the foundation of the kingdom of Kaffa c. AD 1400. It slowly expanded by conquering and subjugating its neighbors; its first city (though not its first royal capital) was Shadda c. 1500 (F. Bieber, 1923:517-518). In the middle of the 16th century it was much harassed by the Oromo (1923:519-520). The king of Kaffa was a king of kings, and bore the imperial title as an equal to the Abyssinian emperor in Gondar from c. 1700. (1920:89) Kaffa's imperial highpoint came in the early decades of the 19th century, when it ruled a state over 100 x 200 miles in extent. (1923:527, map XXII) In 1881 the rulers of Shoa, who later took over the Abyssinian imperial throne, began to invade Kaffa, which they conquered and subjugated in 1897. (1923:532-533; cf. 1920:81-120) From the Abyssinian perspective matters are more complex. Emigrants from the Ethiopian empire had founded a number of states in the south, Kaffa the most important, sometimes under the empire's domination and tributary. (Haberland, 1992:738-749) Ajayi and Crowder's maps show no "Kafa" AD 1200 or 1300; Kaffa well to the south of its present location, outside the "Christian Empire" of Abyssinia under Zar'a Ya'qob AD 1450-1500. (1985:28) Haberland's maps show Kaffa just outside Abyssinia's sphere in 1500 (1992:711). Ajayi and Crowder show Kaffa, plus a shield-perimeter of other Cushite states--Sarka, Enarya, Guraga, Hadya, Janjero, Wollamo, Bali, Sidama--as "areas under less definite Imperial control" in 1527, and outside such control in 1540, with the lands disputed between Abyssinia and the Muslim Sultanate of Adal overlying some of the perimeter. (1985:39) Haberland shows Kaffa as a dependency of Abyssinia in 1550 (1992:708); Ajayi and Crowder show Kaffa adjacent to but independent of Adal and Abyssinia in 1560, though once again those states encroached on the Cushite perimeter. (1985:39) Oromo/Galla from the south expanded against Cush, Abyssinia and Adal after 1527. Abir contends that Abyssinian emperor Serse- Dingel/Sartsa-Dengel (1562-1597) tried to keep the Cushite/Sidama southwest, including Kaffa, under his authority, despite the Oromo invasions which tended to cut all connections, because of its value as a luxury-export-producing area. (Abir, 1975:561-562) He was able by "massive campaigns" to force Kaffa and neighboring kindred states to "become more closely linked" to Abyssinian Christianity; he also treated them as tributaries. (Haberland, 1992:727) But by 1600, Ajayi and Crowder show the Cushite states more or less cut off from both Abyssinia and the Muslim states by the Oromo expansion. (1985:39) Emissaries of Sisinnius (Susenyos) in 1614 could not get through to Kaffa and "marched for several days through a desolate country infested by roving tribes" even to reach "Enarya," where they were turned aside. (Jones and Monroe, 1935:104) Ethiopia was able to force occasional armies through the Oromos to collect tribute; abandoned in the late 17th century, the practice was resumed in 1704. Thereafter however the Oromo further penetrated the Sidama area. (Abir, 1975:561-562) Inariya, one of the Cushite states linking Abyssinia to Kaffa, was conquered by the Oromo in 1710. (Haberland, 1992:728) By the Abyssinian warlord era of 1769-1855 (otherwise "era of the Princes," "age of the Judges"), "Ennaria" and Kaffa had "completely severed their relation with the central government." (Gabre-Sellassie, 1975:2) At odds with Bieber's view that Emperor Hotti Gaocho lorded it over his Oromo neighbors, Ajayi and Crowder show all the Cushite states including Kaffa as "Oromo areas post 1600" on their map for 1800 (1985:39); but Pankhurst and Cassanelli declare that in the early 19th century Shoa reopened the caravan route to Kaffa, which remained independent. (1989:382, 411) As of 1850, Ajayi and Crowder show Kaffa separated from the Christian states into which Abyssinia had dissolved, Shoa ("Shawa") being nearest, by other independent kingdoms and blank areas; invaded about then by Tewodros, Kaffa vanishes from their maps for 1868, 1889, 1890, 1896, but its blank area is conquered by Menelik in 1894 and 1897- 98. (1985:48) Contrary views are expressed in Gabre-Sellassie: Kaffa fought Gojjam in the 1870's, was under Emperor Yohannes' peaceful suzerainty in 1879, was a rebellious kingdom targeted for invasion in 1881. (1975:102, 104, 119) Analysis. The sources are rife with contradictions. The most optimistic dating of an autonomous Kaffan civilization is c. 1450- 1880, and assumes that the chronicles of Kaffa omit the suzerainty of Abyssinia because there was none, that those of Abyssinia misrepresent sporadic armed caravan trade as tribute, and that Kaffa retained a city even under the harshest Oromo pressure c. 1800. The least optimistic dating (never) believes in Kaffa as an Abyssinian tributary colony that lost its cities when it lost politicomilitary contact with the metropole during the 18th century disintegration of the Abyssinian state, regaining them again, if at all, only when the Abyssinian empire regrouped and restored military connections in the 19th century. Better data must evidently be sought on the actual relations of Kaffa with Gondar and Shoa. Shadda, Bonga and the other Kaffan capitals merit systematic archaeological exploration; likewise Inariya and the other Cushite bridgelands. 5. NORTHWEST AFRICA No Northwest African cities appear on Chandler's list through 650 BC. Carthage appears on the 430 BC list with 35,000; the lower bound is 18,000. In 200 BC, Carthage appears at 150,000, Cirta (Constantine) at 40-38,000; by then, however, Carthage has been in (and defeated in) in chronic war with Italy, hence is part of the same civilization. In AD 100, Carthage, Cirta, Hadrumetum, Tingis, Volubilis and Rusicada all appear; but by then Northwest Africa is part of the Roman empire. The 430 BC data imply a possible North African/Punic civilization. The nonoccurrence of Punic cities on the 1000 BC, 800 BC and 650 lists would not disconfirm a Punic civilization, given the high lower bounds of these lists. The traditional date for the founding of Carthage is 814 BC. Utica was supposedly earlier. But Warmington places permanent settlement dates on the coast of the Maghrib only after 800 BC, with such settlements at first "perhaps no more than a few hundred settlers at most." (Warmington, 1981:443-444) R.C.C. Law is almost as skeptical of settlement dates before 800 BC. (Law, 1978b:117-118) Warmington finds the Punic settlements in North Africa initially "politically subordinate to Tyre." (1981:443-444) Assyria acquired hegemony in Phoenicia c. 738-627 BC; after a brief reassertion of independence, Phoenicia went under successive Chaldean, Persian and Macedonian hegemonies from c. 574 BC. Greek cities began colonization of Sicily after c. 735 BC, Syracuse beingthe largest of the resultant colonial city-states. Warmington dates the emergence of Carthage as an independent city to the sixth century BC, and attributes it to the weakening of Tyrian power, the subjection of Phoenicia to Babylon, and the attack in 580 BC of Selinus and other Greek cities on Phoenician settlements in Sicily at Motya and Palermo, which Carthage apparently took the lead in repelling. (Warmington, 1981:444) On the one hand, Carthage took independent action earlier, resisting the foundation of Massilia (Marseilles) by the Phocaeans c. 600 BC (Law, 1978b:120); on the other hand, it showed subordination to Tyre much later, sending a share of its revenues to the temple of Melqart in Tyre even to the second century BC (Law, 1978b:119). Under "Malchus" and Mago Carthage fought the Greeks in Sicily and Sardinia from c. 550 BC. (Law, 1978b:120) A Carthaginian- Etruscan alliance won a notable victory in the naval battle of Alalia against the Phocaean Greeks in 535 BC. Analysis. The Punic cities of northwest Africa passed directly and even with some overlap from a period of political subordination to Tyre to a period of military struggle with the Greek city-states. There was no isolated period, hence no autonomous Punic or Northwest African civilization. The Punic cities (and their Greek counterparts and rivals) were the westward extension and semiperiphery of Central civilization. 6. WEST AFRICA (WESTERN SUDAN) West African/Western Sudanic cities. None appear on Chandler's list through AD 622. Thereafter the picture (with sporadic supportive or alternative readings from other workers) is: Size Lower limit AD 800 Gao 72,000 Ghana 30,000 Awdaghost no estimate none Sijilmessa (Morocco) at 24,000 may also be mentioned as the nearest trans-Saharan city to West African civilization on Chandler's city lists. Sijilmessa became an important political and economic center around AD 780. (Lewicki, 1988:280) It would be reasonable to suppose that this was due to a trans-Saharan trade, and therefore that at least one similar prosperous sub-Saharan urban terminus must have arisen no later than this time. Songhay began trans-Saharan trade with Tahert, Algeria in the mid- 8th century, and developed the trade town of Gao north of the old capital Kukiya. (Levtzion, 1978:677-678) AD 900 Gao 40-36,000 30,000 Sijilmessa appears with 35-32,000. AD 1000 Ghana 30,000 Tademekka none none Sijilmessa appears with no estimate. Awdaghost was "a big town," "a town of several thousand inhabitants, very busy in the tenth and eleventh centuries and undoubtedly struck by a disaster in the middle of the latter century." (Devisse, 1988:414-416) AD 1100 Gao 35-32,000 30,000 Songhay moved its capital to Gao by the mid-10th century. (Levtzion, 1978:677-678; Niane says toward the end of the 11th century, 1984:123; Cissoko says in the 12th century, 1984:187-188) AD 1200 Gao 25,000 + Ghana 25,000 Walata 25-20,000 Zagha 20,000 none Sijilmessa also appears with no size estimate. Takrur city (site not located; Chandler's Zagha) was a metropolis according to al- Bakri. (Niane, 1984:123) AD 1300 Mali 40,000 Walata 25-20,000 Gao 25,000 Jenne 20,000 none Probably during the reign of Mansa Musa I (1307-1332), Walata started to become important, and Jenne and Timbuktu "started to develop into cities that were to be world-famous a century later." (Niane, 1984:151) AD 1400 Mali 50,000 Gao 40,000 Walkokh 30-20,000 Jenne 20,000 Timbuktu 20,000 Ouagadougou none none "The [Mali] capital, Niani, had at least 100,000 inhabitants in the fourteenth century" (Niane, 1984:156) Walkokh (Jolof), in Senegal, had its connections with the Upper Niger. From the second half of the fourteenth century, Timbuktu replaced Walata as the chief southern caravan terminus. (Levtzion, 1977:416) Ouagadougou (of the Mossi), though not on the Niger, is near enough to Jenne, Mali (and Gao) for it to be most unlikely to have constituted a new civilization. Mossi politicomilitary relations at this period were with the Middle Niger. AD 1500 Gao 60,000 Mali 40,000 Walkokh 25-20,000 Jenne 20,000 Ouagadougou none none Niane gives Mali/Niani some 60,000 inhabitants at the start of the sixteenth century. (Niane, 1984:156) AD 1600 Timbuktu 25,000 none Cissoko gives figures of 30-40,000 for Jenne, 80,000 for Timbuktu (vs. Mauny's 25,000), and nearly 100,000 for Gao--the "three largest towns"--toward the end of the 16th century. (Cissoko, 1984:206) Presumably these figures predate the catastrophic demise of the Songhay empire, which Chandler's reflect. AD 1700 none none AD 1800 Segu 30,000 none Earlier datings for a West African civilizational startup? There are suggestions that there was a state of Ghana as early as the 10th century BC or at all events in the third century AD. (De Medeiros, 1988:129) "Community life" or "town life," with trade, division of labor, and metal-working was to be found sub-Sahara starting as early as the second century AD. (Devisse, 1988:376) There was trans-Saharan trade with Roman and Byzantine North Africa. (Andah, 1981:618) Niane asserts that Jenne-Jeno (Old Jenne) was settled in the third century BC, and at its greatest extent, some 34 hectares, around AD 500. (Niane, 1984:117-118) Niane sees new Jenne as a major commercial center by about AD 800. (Niane, 1984:118) A "town" existed at Jenne-Jeno between 400 and 900; it "developed greatly" from 900 to 1400. (Devisse, 1988:417) Chandler estimates it at 20,000 by 1400. Evidently more digging will be needed to locate the crossing of the demographic threshold, and to decide whether Kumbi Saleh, Awdaghost, Kugha and Jenne grew by accretion or by sudden condensation. West Africa in the 9th century AD. There were complex states as well as cities. Ghana was certainly an empire, i.e. a kingdom with a king-of-kings, by no later than AD 872. (Levtzion, 1978:667) Levtzion notes a gift-deputation from the ruler of Tahert, Algeria, to the "king of the Sudan" about AD 871, but attributes no political substance to it. The question of whether the Upper Niger and Middle Niger areas formed a single sociopolitical network is worth raising. Apparently they did so, and quite early. Already by the 9th century, Tademekka dominated an area to its south, opposite Gao on the Middle Niger (9 days walk away), secured gold from Ghana (Upper Niger area), and warred with it. (Lewicki, 1988:306) West Africa in the 10th century AD. By the middle of the next century, the far western and far eastern ends of the system were in touch: apparently around AD 960, the king of Awdaghost exchanged gifts with the ruler of Kugha (i.e. old Gao). (Levtzion, 1978:671) These gifts had political meaning: Kugha "was at that period so powerful that the King of Audaghost thought it prudent to make presents to the king of that place (the King of Songhay), in order to prevent him from making war upon him." (Barth, 1965:III, 658) Sijilmassa and Awdaghost were the northern and southern termini of the major western trans-Saharan commercial routes. But trans-Saharan political bonding well postdates this trade, leaving a window for the autonomous development of at least one sub-Saharan civilization. There was plenty of politicomilitary activity at both ends of the route. In the tenth century, Fatimids, Umayyads, and their local allies fought over Sijilmassa. In the tenth and eleventh centuries sedentary Ghana and the nomadic Sanhaja contested for Awdaghost. But between Sijilmassa and Awdaghost control of the routes belonged to nomads. (Levtzion, 1978:647-652) Devisse argues cogently that the trans-Saharan route from Tahert was open from the second half of the eighth century, but infirm until the tenth. (Devisse, 1988:370-389) Sijilmassa was two months from Awdaghost according to Ibn Hawkal (c. 950); the latter was ten days' march from Ghana city. The tenth century writer al-Muhallabi gave forty days rather than two months, through sand and desert. (Lewicki, 1988:310) Absent reports to the contrary, it is fair to assume that this journey prevented significant politico-military interaction, even though the desert was notoriously permeable by trade. West Africa in the 11th century AD. When was the intercivilizational politico-military barrier finally breached? With the Almoravids? Levtzion believes that "the Almoravids' exploits marked a decisive stage in a long process in which black sedentaries retreated south to the Sahel as the Berber nomads advanced." (Levtzion, 1978:665) For nomads to displace a civilization (because a great drought made its way of life untenable) (Levtzion, 1978:665-666) surely implies that it is less integrated, rather than more so, with neighboring civilizations. Consistent with Levtzion's argument is the fact that rather than controlling Awdaghost the Almoravids destroyed it in 1055. (De Medeiros, 1988:135) Their alleged justification was its submission to Ghana. (Lewicki, 1988:311) From our perspective, what is significant is that (a) they could destroy it at all--it was weakened; and (b) it was not refounded--its viability must have been marginal already. Levtzion nonetheless asserts that with the Almoravid conquest "the western Sudan...became more closely attached to the Maghrib" far beyond the period of the occupation, after having "previously been connected with the Maghrib by enterprising traders only." (Levtzion, 1977:331) But the connections cited are economic--more trade--and cultural--more proselytization--rather than politicomilitary. As regards the politicomilitary connection, the victorious Almoravids found it both necessary and impossible to be on both sides of the Sahara at once. They had to cross and recross between northern and southern termini in the 1050's, and then to divide into northern and southern wings in the 1070's and 1080's, to conquer both Morocco and Ghana. (Levtzion, 1977:333-335) In 1083 the Almoravids and Ghana enforced Islamic orthodoxy in Tademekka. (Levtzion, 1977:349) But Almoravid political domination in the Sudan was brief, perhaps only a decade, and left but a few linkage- traces. (Levtzion, 1977:335-336, 349) The Almoravid legacy was also one of intracivilizational political dissolution rather than integration. After the Almoravid conquest, Ghana became but one of several states: Diafunu to its west, Mema to its east, Susu to its south (Levtzion, 1977:351), Tekrur further west extending its authority up the Senegal to the former sphere of Ghana. (Levtzion, 1977:353) Again, this was a continuation rather than a reversal of previous trends: the provinces of Mande (Manding) and Takrur had thrown off Ghana's rule "as early as the middle of the eleventh century." (Niane, 1984:119) West Africa in the 12th century AD. Takrur reached its zenith between the end of the eleventh and the middle of the twelfth century, controlling the city of Barissa and the salt-mines of Aoulil. (Niane, 1984:120) Soso (its city unidentified) expanded toward the end of the twelfth century. Its king Sumaguru Kante fought with Mande nine times, ruled all the ex-Ghana provinces except Mande and made that a vassal, driving it into revolt. (Niane, 1984:125-126) This is not a well-documented century. Between the fall of Kumbi Saleh to the Almoravids around 1076 and the victory of Sundiata in 1235, little is written about the Western Sudan. (Niane, 1984:118) West Africa in the 13th century AD. Sundiata of Mande, with a force from Mema, led the Malinke revolt against Soso and razed the town of Soso. He conquered most of what Ghana had controlled. Jolof, Diafunu, Takrur and Gao were all conquered by him or his immediate successors between 1220 and about 1310. (Niane, 1984:130-147) Takrur, contemporary of Ghana, had been independent of it, but fell within the sphere of Sundiata Mali (c. 1230-1260). (Fage, 1977:484-486) In the 13th century, the Mali empire "seems to have imposed its hegemony over the whole of Senegambia" for a century. (Person, 1984:313) West Africa in the 14th century AD. Mali conquered Songhay by about 1300; Songhay however shrugged off Mali's yoke by the end of that century. (Niane, 1984:147, 161) Mali under Mansa Musa (1312- 1337?) seems to have had authority over the emerging Jolof kingdom of the Wolof people of the lower Senegal. Until about the second half of the 14th century the Wolof remained within Mali's sphere of influence, and then became independent under the Jolof "empire." (Levtzion, 1977:381,456) The Jolof "confederacy" had formed after 1360, and became hegemonic over northern Senegambia (Barry, 1992:263), taking over Takrur, further inland. (Fage, 1977:484- 486) There were tentative trans-Saharan politicomilitary interactions in this century, which promised more than occurred. Princes of the Maghrib asked Mansa Musa for help in recovering their thrones. (Niane, 1984:164) Apparently they got none. After the Marinid Moroccan conquest of Tlemcen in 1337 brought the main western-route north-end caravan termini under one authority, Mali and Morocco were in diplomatic touch until 1360, but the decline of both states broke off this potential linkage. (Levtzion, 1977:357,395-396) For about two centuries after 1360 the dynastic struggles in Morocco kept it out of politico-military touch with the Sudan. (Levtzion, 1977:410) Succession crises also began in Mali about 1387. (Ly-tall, 1984:174) Under such circumstances, serious connections could hardly have been maintained. Ibn Battuta recounted the difficulties of the journey from Sijilmassa to Walata in 1352. Without guides, the routes could not be found; water for ten days had to be carried; a water party had to come four days' travel out from Walata to reach the caravan. (Levtzion, 1977:370- 371) Sustained politico-military relations would evidently have required coordinated commitments on both ends of the route. West Africa in the 15th century AD. Toward the end of the fourteenth century, Songhay became independent of Mali, and began to expand westward, conquering Mema early in the 15th century. (Levtzion, 1977:420-421) The Tuareg captured Timbuktu, Walata, Nema and perhaps Gao around 1433. (Ly-tall, 1984:174) Called in against the Tuareg about 1468, Sonni Ali of Songhay (1464-1492) took Timbuktu and appointed a governor. Sonni Ali then expanded into the upper Niger. He conquered Massina and, in 1473, Jenne; he drove out a Mossi invasion in 1483. (Levtzion, 1977:420-426; Cissoko, 1984:191) A Fulani state in the Futa Jallon broke away from Mali control about 1490, conquered Futa Toro, and fought Mali till 1514, reducing the latter's sphere of influence. (Levtzion, 1977:457-458. Niane, 1984:180-181) Tentative politicomilitary nibblings at the fringe occurred toward then end of the century, presaging the larger-scale intrusions of the next. In the 15th century European explorers found five vassal kingdoms of the Jolof empire (whose capital was inland) upon the Senegal-Gambia coast. (Niane, 1984:130-147) In 1455 the Portuguese established commercial relations with the Wolof south of the Senegal river. (Levtzion, 1977:452) By the 1480's the Portuguese were using state power on the coast and the seas to monopolize the Senegal sea trade. (Fage, 1977:503-504) In the 1480's Portugal intervened in internal struggles in order to support a candidate for the Jolof empire, which broke apart instead. (Levtzion, 1977:456-457. Fage, 1977:508) The Portuguese sent embassies to Futa, Timbuktu and Mali between 1481 and 1495. (Ly-tall, 1984:182-183) Portugal and Mali established diplomatic relations from "the fifteenth century onwards." (Ly-tall, 1984:174. Fage however refers to attempts to establish diplomatic relations with Mali as unsuccessful; 1977:508) The Portuguese trade assisted small coastal kingdoms--Waalo, Bawol, Kayar, Siin, Salum-- in freeing themselves from Jolof. (Barry, 1992:267-268. Ly-tall mentions Mali rather than Jolof; 1984:182-183) West Africa in the 16th century AD. Askiya Muhammad of Songhay (1492-1528) controlled Teghaza in the north and Hausaland in the east. (Cissoko, 1984:194) Adamu (1984:280-281) questions the latter assertion despite the particularity of Leo Africanus' account. Even so, however the argument over control is resolved, politicomilitary contact was clearly established (via Kebbi) with whatever civilizational network extended over Hausaland (q.v., infra.) About 1501 or 1502 Zara (Diara) fell to Songhay under Askiya Muhammad. The Fulani Futa Toro state attacked Zara around 1512 but were driven out by Songhay forces, but then attacked Jolof, expanded against it, and dominated it till the first half of the eighteenth century. Futa Toro attacked Mali's Bambuk gold fields about 1534 and controlled them by 1600. (Ly-tall, 1984:181-182; Levtzion, 1977:457-458) In 1537, Kaabu, the western province of Mali, became independent, gradually evolving into the local empire of southern Senegambia. (Wondji, 1992:389) Kaabu survived as a state until 1867. (Person, 1984:313-315) In 1537 diplomacy again crossed the Sahara. The Sultan of Marrakesh demanded that Songhay hand over the salt mines of Taghaza, but was rebuffed. (Levtzion, 1977:410) In the 1540's the Moroccan sultan sought to control Wadan and Taghaza, but "was unable to employ military force in the desert." (Levtzion, 1977:411) The Moroccans seized Taghaza in 1556-1557, but agreed to share salt revenues with Askiya Dawud of Songhay. (Abitbol, 1992:300-301) Both the Portuguese and the Ottoman Turks showed an interest in the middle Niger about this time, the Portuguese trying in 1565 to reach Timbuktu via Senegal, while the Turks sent expeditions into the Sahara from Algeria in 1552 and 1578-1579, and conquered the Fezzan in 1557. (Abitbol, 1992:301) The Portuguese were detained by struggles involving Bainuk and Kasanga interests on the Senegambian coast, which had organized themselves sufficiently to fight c. 1570-1590. (Barry, 1992:266-267) By the 1580's Morocco had become able to extend itself militarily in the Sahara. The Moroccan caliph al-Mansur seized Saharan Algeria in 1582, accepted an alliance with Bornu, doubtless against the Turks, in 1583, made an abortive attack on Senegal in 1584, nearly attacked Songhay in 1586. (Abitbol, 1992:301) With Songhay badly weakened by its civil war of 1588, musketeers providing Morocco with military superiority, and the prospect of gold and slaves to be acquired, Morocco finally invaded the Sudan 1590-1591 and defeated Songhay at Tondibi. (Levtzion, 1977:411- 414, 440-442) A puppet Askiya was established in Timbuktu, while an independent resister set himself up at Dendi. (Levtzion, 1977:443) The Moroccans conquered Gao, Kukya, Timbuktu, Jenne and other river ports and destroyed the Songhay empire 1590-1594. (Abitbol, 1992:304) In 1599 Morocco defended Jenne from attacks by Mali. (Levtzion, 1977:455) The role of military technology in establishing politicomilitary connections has here to be acknowledged. In his first, failed attempt on Songhay "the Emperor of Morocco sent a very numerous host, said to be 20,000 strong.... But this time also the danger passed by, the numbers of the army themselves causing its ruin in consequence of hunger and thirst." (Barth, 1965:III, 673-674) The second, successful Moroccan army numbered only 3600--but 3600 musketeers, and "the Songhay do not seem to have possessed a single musket." (1965:III, 675, 676) West Africa in the 17th century AD. By now the penetration and entrainment of the Western Sudan by Central states and processes was well advanced. In the seventeenth century, the Dutch, French and English broke the Portuguese trade monopoly on the Guinea coast. (Barry, 1992:268-271) In 1641, the Dutch compelled the other European states to acknowledge their trading monopoly over the whole Guinea coast. (Wondji,1992:390) But France and England, among others, soon undertook to break that monopoly too; forts and trading posts went up in Senegambia and elsewhere. (Wondji, 1992:390-391) The process of fortifying trading posts meant that during the seventeenth century the Senegambian coast was partitioned into Dutch, French, English and Portuguese spheres of influence. The French eventually got the lion's share. (Barry, 1992:262, 268-271) The sultans of Morocco turned their thrust west from the Niger basin, to counter and rival the Europeans, undertaking to penetrate, subjugate and raid in Mauritania and Senegal. (Abitbol, 1992:314-315) Morocco stopped appointing or reinforcing the pashas of Timbuktu after 1618, and the pashalik developed into an autonomous settler-colonial state, with the Bambara state of Segu and the Tuareg confederacy as its neighbors. (Abitbol, 1992:307) The pashas of Timbuktu generally gave allegiance to the sultans of Morocco, who made few or no demands on them. (Abitbol, 1992:314- 315) By the late seventeenth century the pashalik had devolved into four more or less autonomous garrison states. (Abitbol, 1992:308) A residual Songhay state, with the help of Kebbi, maintained some independence. (Abitbol, 1992:310-312) The disruption caused by the Atlantic slave trade provoked c. 1659-1677 a nominally anti-slavery religious and revolutionary war in the vicinity of the French sphere in Senegal; however, the war was economically based on the trans-Saharan (slave, etc.) trade. In the first phase of the war, Berber marabouts overthrew local aristocracies. The French, preferring to face small, warlike aristocratic states rather than a large Islamic political grouping which might have broken the monopoly and controlled terms of trade on the Senegal river, intervened to restore the fallen aristocracies 1674-1677. (Barry, 1992:274-277) West Africa in the 18th century AD. Through the eighteenth century, the French intervened in succession crises on the Senegambian coast to keep the states there small; they fought the Dutch and English there, with local allies and enemies also involved. Morocco also became militarily entangled in the Senegambian struggle. (Barry, 1992:279-285) The seventeenth century defeat of the Muslim marabout movement on the coast was followed by repression, diaspora, Muslim revolutions, and Muslim theocracies. But even in the inland revolutionary jihad states it was possible for the French to intervene: Futa Toro, the nearest to St. Louis on the Senegal and the most aggressive toward the end of the eighteenth century, found its influence limited by French aid to its rebels, secessionists, and external enemies. (Barry, 1992:295-299) Analysis. There was a historically autonomous civilization in West Africa, from Senegambia through the Middle Niger, which maintained an autonomous politicomilitary process despite brief episodes of linkage to Central civilization via Morocco in the eleventh and fourteenth centuries. Already jeopardized by Portuguese penetration of Senegambia toward the end of the fifteenth century, its autonomy was definitively ended in the sixteenth century by a successful Moroccan invasion of the Niger basin, forestalling attempts at penetration both by Portugal via Senegal and by the Ottoman empire via Algeria. This successful preclusive imperialist venture was followed by a passive exercise of Moroccan suzerainty on the Niger and a shift in the locus of penetration. The 17th and 18th century saw a struggle among Moroccans, Europeans and local states and revolutionaries in Senegambia. 7. CHADLAND Chadian cities. None appear on Chandler's lists through AD 900. Then: Size Lower limit AD 1000 Manan/Matan 40-35,000 none AD 1100 none 30,000 AD 1200 Njimiye 25-20,000 Ngala none none AD 1300 Njimiye 40,000+ Ngala none AD 1400 Krenik none none AD 1500 Kazargamu 40,000 Ngala 25-20,000 none AD 1600 Kazargamu 60,000 Masenya 20,000 none AD 1700 Kazargamu 50,000 Masenya 30,000 none AD 1800 Kazargamu 50,000 Masenya 30,000 none Other data and datings may be noted. The place Kanem, under the Zaghawa, with "no towns," is mentioned by al-Yakubi AD 872. (Lange and Barkindo, 1988:445) Al-Muhallabi (late 10th century) mentions "two towns, Manan and Tarazaki." (Lange and Barkindo, 1988:450) Between the 9th and 10th century the Zaghawa nomads of Kanem had thus acquired two towns, Manan being the capital, and become settled. (Levtzion, 1978:680-681) Al-Idrisi, writing AD 1154, mentions Manan and Njimi as towns. (Lange and Barkindo, 1988:457) In the mid-12th century, Manan was still a small town. (Levtzion, 1978:681-682) Manan had been the Zaghawa permanent capital for at least a century prior to the shift to Djimi about 1100; Djimi was the Sefuwa capital for three centuries. (Lange, 1984:244) Chadland to the 9th century AD. The record is confused and will not be clarified without considerable archaeological work. The So mounds south of Lake Chad show occupation AD 500-1300. There may or may not have been politico-military ties with the Niger delta. (Fage, 1977:473-474) Lange and Barkindo (1988:447- 448) propose that a large trading state of Kanem is already implied in a report of Wahb b. Munabbih (d. AD 730). "The extent of Nubia's relations with Kanem-Bornu...remains uncertain pending further systematic archaeological work. The key area for investigation is Darfur...." (Kropacek, 1984:420) Arkell sees Tundjur rule in Darfur as under Nubian protection, and flourishing between the 8th and 10th centuries. (Kropacek, 1984:420-421) Chadland, 10th century AD. A more conventional view is that Bornu-area states were beginning to form about the 10th and 11th centuries. (Fage, 1977:476-477) Lange and Barkindo assert that there were "diplomatic relations" and "diplomatic missions" between North Africa and Kanem, perhaps from AD 992, certainly from 1228. The examples they cite (gifts of slaves and a giraffe) could however be interpreted as trade-promotion only. (Lange and Barkindo, 1988:452) Chadland, 11th century AD. In the middle of the eleventh century, Arku b. Bulu, Zaghawa king of Kanem, established slave colonies northward into the Fezzan. (Lange and Barkindo, 1988:453) Evidence of reciprocal southward politicization is currently lacking. Lange and Barkindo date the Islamization of Kanem, and the replacement of a Zaghawa by a Sefuwa dynasty, to the last half of the 11th century. (Lange and Barkindo, 1988:454-460) Chadland, 12th century AD. The trade route from Kanem north ran through Kawar (past the salt mines of Bilma and Agram) to the Fezzan and Tripoli. In the first half of the 12th century Kawar and the Fezzan were independent. In 1172-1173 the Mamluks of Egypt conquered the Fezzan. (Lange, 1984:249-252). Chadland, 13th century AD. In the first half of the 13th century Kanem took control of Kawar and the Fezzan. (Lange, 1984:251-252) Lange and Barkindo give 1228 as the date from which there were certainly diplomatic relations between Kanem and North Africa. (1988:452) Arkell sees a conquest of Darfur by Dunama of Kanem c. 1240, with strong influence for the next 400 years, until the establishment of Fur Kayra power about 1640. (Kropacek, 1984:420-421) By the mid-13th century, Kanem was a powerful expanding kingdom. (Levtzion, 1978:681-682) Chadland, 14th century AD. Kanem probably continued in control of Kawar and the Fezzan to the middle of the fourteenth century. (Lange, 1984:254) Toward the end of the fourteenth century the Bulala people conquered Kanem, driving the Sefuwa into the previously vassal state of Bornu to the west. (Lange, 1984:257-260) Bornu made a diplomatic complaint to Egypt in 1391 concerning Arab slave-taking (Kropacek, 1984:422), which could imply that during this period--when Makuria had almost vanished in Nubia--Bornu regarded Egypt as a neighboring state. Chadland, 15th century AD. Kanem and Bornu were both strong states around the end of the fifteenth century, with Kanem in good relations with Egypt. In the late 15th century the city of Gazargamo was built to serve as the Sefuwa capital, as it would for more than 300 years to come. (Lange, 1984:257-260, 265) Gazargamu was built as a fortress in about 1472. (Barkindo, 1992:493) Chadland, 16th century AD. The Bulala of Kanem and Sefuwa of Bornu fought indecisively and sporadically in the first half of the sixteenth century. (Barkindo, 1992:493-494) Around 1564-1577 Alawoma of Borno embarked on a course of local imperialism, conquest, deportation, enslavement and colonization. Kanem was partly conquered and partly subjected by the 1580's. (Barkindo, 1992:496-498) By the early 1500's many small states had appeared south of Borno, among them Bagirmi; some it fought, some it subjugated, some cooperated peacefully. (Barkindo, 1992:494) In the first half of the sixteenth century Borno had conflicts with Kano, perhaps over control of the expanding settlements on the Chad-Hausaland trade route. (Barkindo, 1992:494) Bornu developed a deep military penetration, with a tribute system, in Hausaland in the sixteenth century. (Fisher, 1975:114-115) In the sixteenth century Borno had diplomatic and trade relations with the Maghrib, Tripoli and Egypt. (Barkindo, 1992:496) The Ottoman Turks conquered the Fezzan in 1557. (Abitbol, 1992:301) Borno made several demands on the Ottoman empire around 1574 as regarded security cooperation and the government of the Fezzan, which were conceded in 1577. (Barkindo, 1992:502-503) Nevertheless, in 1582 the king of Bornu sought military assistance from Morocco, possibly against Ottoman incursions, and pledged homage to him. (Levtzion, 1977:414-415) Late in the sixteenth century the subject-allies Mandara to the south and Bagirmi to the southeast began to assert their independence and were attacked by Borno. (Barkindo, 1992:504) Chadland, 17th century AD. The Borno empire was stable in the seventeenth century. Vassals were installed to east and west and buffer states set up along the desert fringe. Wadai was set up beyond Bagirmi, probably to check it, within Borno's political ambit. (Barkindo, 1992:511-512) Wadai was created somewhere in the vicinity of 1611-1635; it emerged in the seventeenth century as the principal power east of Bornu. (Fisher, 1975:136, 139) Around the same time a Darfur state emerged, to become Wadai's suzerain. "Sometimes Wadai paid tribute both to Bornu and to Darfur." (Fisher, 1975:139-140) In the mid-seventeenth century, the Kwararafa pressed on Hausaland, and Bornu helped repel them. (Fisher, 1975:116) In that period, Bornu had diplomatic (gift-exchange) relations with Tripoli. (Fisher, 1975:121-122) Borno, Tripoli and Fezzan cooperated in an unsuccessful attempt to monopolize trans-Saharan trade. (Barkindo, 1992:512) Chadland, 18th century AD. In the early eighteenth century, Bornu marched against Kano, and Zaria joined the ranks of its tributaries. (Fisher, 1975:116) The Kwararafa fought Bornu in the 18th century; by its end they were tributary. (Fisher, 1975:134- 135) In the late eighteenth century, the Borno hegemony withered and Bagirmi, Wadai, Mandara, and Gobir in Hausaland successfully revolted. (Barkindo, 1992:513) Wadai also got free of Darfur in the second half of the eighteenth century, becoming overlord of Kanem, and even subjugating Bagirmi early in the 19th century. (Fisher, 1975:139-141) Though Bornu's influence in Hausaland declined by 1800, the Hausa princes appealed to Bornu to protect them against the jihad of Usman dan Fodio. (Fisher, 1975:66) Analysis. Repeated indications of political contact with Central civilization make the historical autonomy of a Chadland civilization suspect; repeated lapses in the record, or weaknesses in the contacts, render it nonetheless possible. Both the size of Chadian settlement and the degree of Central involvement are at their most ambiguous before AD 1000. Nubia was expansive militarily and diplomatically in the 8th-10th centuries. A colonial, imperial or reaction state forming at Kanem under Nubian pressure through Darfur cannot be ruled out; nor can a So- nomad interface formation. Excavation in Darfur, the So mounds, and Manan (when located) must be sought to shed light on the period; not even a guess can be reasonably ventured. The 11th-14th centuries show alternation in Kanem and Egpyt controlling the Fezzan, and signs of (sporadic? consequential?) diplomatic relations with North Africa. Kanem's 13th-century expansion into Darfur seems concurrent with increasing Egyptian pressure on Nubia; was it coordinated? After the Kanem-Bornu split, Egypt seems to have shifted its connections from the previous to the later rulers of Kanem, i.e. to have maintained state rather than personal relationships, a realpolitik approach which suggests a calculation of advantage. Contacts were particularly strong in the sixteenth century, when Bornu was at its most expansionist. The silent record for the next two centuries most likely indicate that Bornu was a status quo power, and Turkey and Morocco busy expanding or defending elsewhere. A Chadland/Central Sudanic civilization cannot be ruled out. The Darfur and Fezzan connections may prove tenuous, the diplomatic relations inconsequential before the 16th century. I would however treat such a historically autonomous civilization as no more than possible, and expect rather that European ignorance that there was a significant regional power of the Central system in Chad after about the 13th century stemmed from the fact that it was both non- coastal and Muslim, and in no position to be a threat or a target, and not in need of other allies than those locally available to it. 8. HAUSALAND Hausaland cities. None appears on Chandler's list through AD 1100. Then: Size Lower limit AD 1200 Kano 25,000 none Fage dates the formation of Hausaland states to about the 10th or 11th centuries. (Fage, 1977:476-477) Adamu dates the founding of the city of Kano about 1100, its walls completed by 1200, but its hinterland not entirely subdued till c. 1400. (Adamu, 1984:271) AD 1300 Turunku 40-30,000 Kano 30,000 none AD 1400 Kano 30,000 Turunku 30-20,000 Durbi 30-20,000 none AD 1500 Kano 50,000 Turunku 40-30,000 Gobir 28,000 Agades 25-20,000 Katsina 25-20,000 none Agades was a caravan and political center from the fifteenth century. (Levtzion, 1977:433) "Gobir" may have existed at Marandet from the ninth to the fifteenth century, but was reloacted to the present Gobir from the middle of that century, where it protected a salt route. (Adamu, 1984:276) Katsina was established as a walled city in the last half of the fifteenth century and conquered a local kingdom. (Adamu, 1984:273) AD 1600 Zaria 60,000 Katsina 60,000 Kano 40,000 Surame (Kebbi) 40,000 Gobir 35,000 Agades 30,000 none Kebbi was a Songhay sub-province in the late fifteenth century. (Adamu, 1984:277-278) Zaria was established at about the end of the fifteenth century by the merger of two growing town-kingdoms from perhaps the fourteenth century, Turunku and Kufena. (Adamu, 1984:274-275) AD 1700 Katsina 60,000 Zaria 50,000 Agades 45,000 Kano 35,000 Naya/Chibiri 30,000 Kebbi none Zamfara none Puje none none Zamfara may have had a permanent capital from early in the sixteenth century. (Adamu, 1984:276-277) AD 1800 Katsina 70,000 Alkalawa 50,000 Zaria 40,000 Kebbi 30,000 Kano 30,000 Kiawa none none Hausaland in the 14th century. Kano subjugated its hinterland and destroyed its rival city Santolo in the late 14th century. (Adamu, 1984:271) The Jukun empire in the southern Benue basin, whose capital Adamu identifies as Kwararafa, was a powerful state in the fourteenth, perhaps even the thirteenth century. (Adamu, 1984:282-283) In the late 14th century Kano fought the Kwararafa/Jukun and Zaria/Zegzeg, i.e. Turunku and/or Kufena. (Adamu, 1984:272, 274-275) Malian traders built up trade and proselytized in Kano in the second half of the 14th century. (Levtzion, 1977:375) But Adamu avers that Mali "never played any political role in Hausa history." (Adamu, 1984:280) Bornu ties to Hausaland began to develop with the foundation of Bornu shortly before 1400. (Fisher, 1975:114) Hausaland in the 15th century. Hausa-Bornu political relations became critical after 1425, when the deposed Bornu ruler Uthman Kalnama sought refuge in Kano. Bornu regarded this as a threat, and reduced Kano and Katsina, and perhaps all Hausaland, to vassal status. (Adamu, 1984:279) In the middle of the fifteenth century Kano paid tribute to Bornu and began an unsuccessful hundred years' war with Katsina. (Adamu, 1984:272) Laya, 1992:458, says the first Kano-Katsina war broke out in the late fifteenth century and lasted eleven years without result. In the same period Katsina fought Nupe (which had fought the Yoruba as well; Adamu, 1984:274) and Borno fought Jukun (Fisher, 1975:134- 135). Hausaland in the 16th century. Hausaland was polarized between Songhay, later replaced by Kebbi, in the west, and Bornu in the east. Askiya Muhammad of Songhay sent military expeditions to Agades and Hausaland in the early 16th century. (Levtzion, 1977:430) He sought to close Hausaland's trade with Borno. (Laya, 1992:456)He twice attacked Agades, once in 1499/1500, once in 1515, and secured tribute. He is described by Leo Africanus as having conquered Katsina, Zaria and Gobir, depopulating them, and made Kano tributary. (Levtzion, 1977:433.) Adamu doubts that Songhay conquered all Hausaland, as the Hausa chronicles do not mention the event (1984:280-281); Barth believes Leo misattributed the conquest of Hausaland by Muhammadu Kanta of Kebbi, a Songhay vassal who revolted in 1515, to Songhay itself and to Askiya Muhammad. (1965:III, 669-670) Assuming Leo to be correct, however, the 1515 revolt of Songhay's vassal state Kebbi let Hausaland resume independence of Songhay. (Levtzion, 1977:433) Kebbi then built cities, Surame, Birnin Leka and Birnin Kebbi, by merging villages and fortifying the resultant towns against Songhay. (Adamu, 1984:277-278) Kebbi then seized Air/Agades from Songhay, itself conquered Hausaland and parts of Bornu, and invaded Nupe and Yawuri. Bornu counterinvaded the Hausaland vassal states but was defeated. (Adamu, 1984:277- 278) Songhay unsuccessfully attacked Kebbi in 1533 and 1553. (Laya, 1992:456) Borno conquered Agades about 1532. Kebbi raided it and Borno was called on in 1561, with violently indecisive results. (Laya, 1992:456) Kebbi's empire did not much outlast its founder, Muhammadu Kanta, who died in 1556; by the end of the 16th century, Hausaland and Agades were lost to Kebbi. (Adamu, 1984:278) Bornu developed a tributary system in Hausaland (Fisher, 1975:114-115); the Hausa states fell "completely within the Bornu sphere of influence." (Levtzion, 1977:433) Kano, Katsina, Zaria and Jukun/Kwararafa were all active on a more local scale in the sixteenth century. Zaria fought to expand against Nupe and Kwararafa. (Adamu, 1984:274-275) Toward the middle of the sixteenth century, late in its war with Katsina, Kano inflicted a defeat on Zaria. (Adamu, 1984:273) (Laya, 1992:458, sees the second and third Kano-Katsina wars occurring around 1570 and 1580, again without decisive results.) Kwararafa defeated Kano in 1582, causing its people to flee to Daura until 1618. (Laya, 1992:457) When Morocco conquered Songhay in the late sixteenth century, Kebbi sheltered the Songhay resistance despite Moroccan threats. (Laya, 1992:456) In the 1590's, the Sultan of Morocco believed that Kebbi had interfered with the attempt of Katsina and Kano to pay him allegiance. (Levtzion, 1977:415) Hausaland in the 17th century. Kebbi and Bornu were contending great powers. Kano, Katsina, Azbin (Agades and its Air vicinity), Gobir, Jukun, Zamfara and Adar (NW of Gobir) were all active, Nupe and Zaria quiet. A succession struggle in Azbin in the early seventeenth century led to Kebbi and Borno supporting rival claimants, with Kebbi successful. (Laya, 1992:456) There were repeated Kano-Katsina wars in the first half of the seventeenth century; then Kano and Katsina signed a peace treaty c. 1649-1651, which held, out of fear of the Kwararafa. Indeed in 1653, the latter attacked and drove out the ruler of Kano. (Laya, 1992:457, 460) In 1674, with Gobir fighting Kano and Katsina, Zamfara attacked and defeated Kebbi and Adar; Adar then revolted against Kebbi (or was conquered by Borno). (Laya, 1992:456, 462) In 1675, Zamfara defeated Azbin and was in turn defeated by Azbin. (Laya, 1992:456) Borno fought and defeated the Kwararafa in 1680. (Laya, 1992:457) Azbin attacked and defeated Gobir in 1689. (Laya, 1992:456) Hausaland in the 18th century. Kebbi declined, as did Kano. Azbin and Zamfara rose, to be overtaken by Gobir, which finally outclassed Bornu. Zamfara defeated Kano in the early eighteenth century. (Laya, 1992:462) Early in the century, Bornu marched against Kano, and Zaria joined the ranks of Bornu tributaries. (Fisher, 1975:116) Azbin attacked Surame and killed the Kanta of Kebbi in 1721. The Kebbi court moved west in 1722. (Laya, 1992:456-457) Gobir fought Kano repeatedly in the middle of the eighteenth century, with greater and greater advantage. Zamfara intervened against Gobir, but was attacked and severely diminished about 1762. (Laya, 1992:462-463) The Kwararafa fought Bornu in the 18th century; by its end they were tributary. (Fisher,1975:134-135) Toward the end of the century an empire was finally formed out of the city-states of Hausaland, which had previously all managed to survive their 16th-18th century warfare. The imperial capital was Sokoto; the empire extended to the Bornu frontier and penetrating Yorubaland; the method was holy war; the founder was Usuman dan Fodio, fl. c. 1774-1817. (Hiskett, 1976:134-141) Gobir had successfully revolted against Bornu (Barkindo, 1992:513); it was to Bornu that the Hausa princes appealed against the jihad of Usman dan Fodio. (Fisher, 1975:66) Analysis. The enormous amount of local city-state interaction does not conceal the fact that Hausaland was strongly connected politically and militarily to Chadland from at least the 15th century, and that such a connection appeared almost as soon as Kano began to assume more than very local political significance. Connections to the lower Niger appeared late in the 15th century; those with the Middle Niger (and Morocco) came in the 16th century, but seem less salient. The question of the urban status and history of Kano is important. If it was a city AD 1200 (Chandler) or 1100 (Adamu) and had not even peaceful political relations with Chadland prior to the struggle with Bornu in the 15th century, or with the Middle Niger prior to the struggle with Songhay in the 16th, then a Hausaland civilization (or perhaps a Nigerian civilization--cf. Lower Niger, infra--existed for several centuries, till engulfed from Chadland. But the sudden florescence of cities in the 15th century, the assertion that Kano did not even control its own hinterland until the late 14th, and the concern with which Bornu reacted to the flight of Uthman Kalnama, would be consistent with treating Kano as developing from a market town to a city with politico-military significance very late in the 14th century, in the semiperiphery and under the watchful eye of Bornu. Finally, an earlier dating for the city, but also for calculating political relations with Chadland, would place Hausaland in the Bornu semiperiphery, within a system of states whose core power was stimulated to imperialism only when semiperipheral Kano began to grow and to behave assertively (by harboring Bornu's enemies). The third alternative currently seems in accord with the preponderance of evidence: Hausaland was an extension of the civilization previously established in Chadland. A question for future research is how Hausaland's growth is connected to the decline of Western Sudanic cities, whether this is a trade route shift driven by economic developments or the result of political assertion of control "upstream" on a trade route. The latter would imply that the Songhay invasion of the 16th century was a response to a mercantilist assault; that the trans-Sudanic trade routes were relatively more important than has commonly been thought; and that perhaps that Hausaland had to the 15th century been a shared semiperiphery of the Middle Niger and Chadland cities and powers. The archaeology of Kebbi and Zamfara will be important to testing these propositions. Hausaland problems are particularly challenging in that the area had "bridge" states and cities in five directions. To the west, Kebbi and Zamfara linked it to the Middle Niger. To the north, Agades was a bridge to the Mediterranean. To the south, Zegzeg/Zazzau (Turunku, Zaria) and Nupe were bridges to Yorubaland. To the north of Kano and east of Katsina, Daura was a bridge eastward to Chadland. (Jukun to the southeast fought, i.e. was a politico-military bridge to, but is not currently cited as an economic bridge to, Chadland.) Dates of urbanization and contact- politicization in all directions are as important to establishing Hausaland's civilizational status as the dating of events at and around Kano. Chandler's dates would imply that the eastward connection to Chadland came first, in the 12th century; then the southeastward connection (via Turunku and Jukun), in the 13th century; then the southward (via Turunku, Nupe and Oyo), in the 13th and 14th centuries; then the northward (via Durbi, Katsina, Gobir, Agades) in the 14th and 15th; finally the westward (via Surame) in the 16th. The eastward dating is much earlier than Adamu's discussion of politico-economic ties to Bornu in the early 15th century (1984:272, 279). The southeastward dating is consistent with Adamu's discussion of Jukun and Zaria (1984:272, 274-275, 282-283). The southward dating is much earlier than Adamu`s (1984:283-284) or Nadel's (1942:75) 15th-century dating of Nupe. The northward dating is consistent with Adamu's dating of Katsina and Gobir (1984:273, 276) and Levtzion's of Agades (1977:433). The westward dating is later than Adamu's for Kebbi (1984:277-278). I suspect that an E-SE-S-W-N ordering will ultimately emerge, implying that urbanization moved from Chadland to Hausaland and Jukun, to the Lower Niger, and that urban bridges were then built to already-urbanized the Middle Niger (along established and new trans-Sudanic routes) and then the Mediterranean (along the more challenging trans-Saharan route). Politicization was probably concurrent, since any adjacent bridge city was immediately perceivable as a potential ally, looter, lootee, vassal and suzerain. 9. LOWER NIGER Lower Niger cities. None appear on Chandler's list through AD 1200. Then: Size Lower limit AD 1200 Bussa none none Bussa, in Borgu, is a possible bridge city to Hausaland, or to the Middle Niger. AD 1300 Ife none Bussa none Nupe none none Nupe would also be a bridge to Hausaland. AD 1400 Oyo 50,000 Nupe 50-30,000 Ife none Benin none none Adamu dates the Nupe state in the Niger-Benue confluence to the fifteenth or sixteenth century. (1984:283-284) AD 1500 Oyo 60,000 Nupe none Ife none Benin none Ijebu none none AD 1600 Oyo 60,000 Benin 50,000 Gbara (Nupe) 33,000 Ife none none AD 1700 Oyo 50,000 Jima (Nupe) 35,000 Benin 25,000 none AD 1800 Oyo 80,000 Ilorin 40-30,000 Gbogun 40-30,000 Ife 40-30,000 Ogbomosho 40-30,000 Kiama 30,000 Gbara 24,000 none Alternative datings exist. Ife: Fage dates Ife at its site to the 10th or 11th century. (Fage, 1977:476) Andah and Anquandah date a cultural and political ascendancy of "Ife town" in Yorubaland between the seventh and eleventh centuries. (Andah and Anquandah, 1988:513) Benin: Fage dates Benin as a flourishing city by AD 1200-1300. (Fage, 1977:476) For Ryder, Benin became a city (no size estimate), under Ewuare in the fifteenth century. (Ryder, 1984:352-353) "Benin City" of Edo may have had a city wall by the eleventh century and certainly had one by the mid-fifteenth. There the "true urban unit" was to be dated to Oba Ewuare in the fifteenth century; prior to that there was a "town," a fusion of scattered villages. (Andah and Anquandah, 1988:514, 516) Igbo-Ukwu: Igbo-Ukwu is an Iboland urban complex whose dating presents problems, with arguments for ninth to sixteenth-century dates available. (Ryder, 1984:361-366) Igbo-Ukwu, a "town" or "small settlement" with a state, dates from the ninth century. (Andah and Anquandah, 1988:517-520) Cross River: there may have been an Akwanshi centralized state in the Cross River Valley sometime after the 7th century AD and before the onset of the trans-Atlantic slave trade. (Andah and Anquandah, 1988:524-526) In the eastern Niger delta, city-states were established by the eighteenth century in Bonny, New Calabar, Okrika and Nembe, based upon the slave trade and the wealth thereby gained. (Alagoa, 1992:447) In general Andah and Anquandah refer to the urban centers of Guinea in the seventh to eleventh centuries as "towns," though with complex specialization and trade. (Andah and Anquandah, 1988:526- 527) Lower Niger through the 15th century AD. Centralized states in Yoruba, Benin, Nupe, Igala, and Kwararafa/Jukun may have begun to form about the 10th and 11th centuries. (Fage:1977:476-477) Nupe and Igala (Idah) had developed state systems before the 15th century. (Fage, 1977:496, 512) Benin became a large kingdom, conquering some Edo, Ibo and Yoruba peoples, under Ewuare in the fifteenth century. (Ryder, 1984:352-353) At the end of the 15th century, Benin dominated Yorubaland. (Fage 1977:497) In the fifteenth century, Borgu, Idah and Kwararafa were "major states" north of Yorubaland. (Ryder, 1984:351) Alagoa gives an early date for the rise of Oyo: it began after trouble with Nupe in the 15th century. (Alagoa, 1992:442) Lower Niger in the 16th century AD. Fage gives a later dating: Nupe was continually at war with the Yorubas from the early 16th century, invading Oyo around 1535; the Oyo kings fled to Borgu, and likely received Borgu aid in their later return, c. 1610. (Fage, 1977:502-503) For Alagoa, on the contrary, Oyo pushed back Nupe with Borgu help by the beginning of the sixteenth century, and then began to expand against Borgu too. (Alagoa, 1992:442) Another dating is more secure. The Portuguese, seeking to monopolize the sea trade, by the start of the 16th century tried to ally themselves with the administration of Benin. (Fage, 1977:505) They could not however deal with Benin from a position of political strength, though they missed a chance to arm the Oba in 1514. Instead, they opened trade with, and provided a political counterweight to, Benin's smaller, weaker eastern neighbor Warri. (Fage, 1977:515-518; Itsekiri, capital Ode Itsekiri, says Alagoa, 1992:445) Lower Niger in the 17th-18th century AD. This is generally agreed to have been a period of Oyo empire in Yorubaland. Ryder judges Oyo to have "developed its 'imperial' character quite late, perhaps in the early seventeenth century," perhaps after struggling with Borgu and Nupe. (Ryder, 1984:348) Oyo expansion stopped at Ijebu and Benin to south and southeast, but reached southwest as far as Dahomey in the 18th century. (Alagoa, 1992:442) Oyo had strong influence and involvement in the states for one or two hundred miles to its south and east, and complementary spheres of influence with the Benin empire beyond. (Asiwaju, 1989:704-706) The Oyo empire began to fragment in the late eighteenth century; defeated by Egba c. 1774, Borgu 1784, Nupe 1791, the empire collapsed about 1835. (Asiwaju, 1989:706) Lower Niger in the 19th century AD. The collapse of the Oyo empire was followed by Hausa intervention from Sokoto against Borgu and Oyo in 1836 and 1840, by three post-imperial intra-Yoruba wars between 1820 and 1893, and by the independence--and intervention-- of Dahomey from the early 1820's to the French conquest of 1892. (Asiwaju, 1989:707-709) The British acquired Lagos in 1861, the French Porto Novo in 1863. The British-French struggle to control trade, to end slavery, to pacify the interior and to outstrip one another led in due course to the French conquest of Dahomey in 1892, the British conquest of Benin in 1897, and the partitioning of Borgu in 1895. (Asiwaju, 1989:717) Analysis. Was Benin a city in the eleventh century, or the fifteenth? If in the eleventh, it becomes possible that there existed a Lower Niger civilization separated from events in Hausaland. More likely, the rise to urban status above and below the Niger-Benue confluence was simultaneous, and a system of states arose in what is now Nigeria, with Borgu, Nupe, Igala and Jukun playing key intermediary roles. Chronology is problematic, for the area's cities, states and empires alike. The datings of the sizes of Ife, Benin, and the towns in Nupe and Jukun are especially critical here. If Ife and Benin far predate the Nupe and Jukun evolution of cities, then such a Lower Niger civilization was a reality, lasting until fusional bonds were established through Nupe, Jukun, Borgu, Igala, to Hausaland--and, less significantly, through Benin to Portugal--by the 16th century. If Benin did not become a city until the 15th century, and even postdates urbanization in Nupe, then this area is a semiperipheral pressu